xAI Dev Leaks API Key for Private SpaceX, Tesla LLMs

An employee at Elon Musk’s artificial intelligence company xAI leaked a private key on GitHub that for the past two months could have allowed anyone to query private xAI large language models (LLMs) which appear to have been custom made for working with internal data from Musk’s companies, including SpaceX, Tesla and Twitter/X, KrebsOnSecurity has learned.

Philippe Caturegli, “chief hacking officer” at the security consultancy Seralys, was the first to publicize the leak of credentials for an x.ai application programming interface (API) exposed in the GitHub code repository of a technical staff member at xAI.

Caturegli’s post on LinkedIn caught the attention of researchers at GitGuardian, a company that specializes in detecting and remediating exposed secrets in public and proprietary environments. GitGuardian’s systems constantly scan GitHub and other code repositories for exposed API keys, and fire off automated alerts to affected users.

GitGuardian’s Eric Fourrier told KrebsOnSecurity the exposed API key had access to several unreleased models of Grok, the AI chatbot developed by xAI. In total, GitGuardian found the key had access to at least 60 fine-tuned and private LLMs.

“The credentials can be used to access the X.ai API with the identity of the user,” GitGuardian wrote in an email explaining their findings to xAI. “The associated account not only has access to public Grok models (grok-2-1212, etc) but also to what appears to be unreleased (grok-2.5V), development (research-grok-2p5v-1018), and private models (tweet-rejector, grok-spacex-2024-11-04).”

Fourrier found GitGuardian had alerted the xAI employee about the exposed API key nearly two months ago — on March 2. But as of April 30, when GitGuardian directly alerted xAI’s security team to the exposure, the key was still valid and usable. xAI told GitGuardian to report the matter through its bug bounty program at HackerOne, but just a few hours later the repository containing the API key was removed from GitHub.

“It looks like some of these internal LLMs were fine-tuned on SpaceX data, and some were fine-tuned with Tesla data,” Fourrier said. “I definitely don’t think a Grok model that’s fine-tuned on SpaceX data is intended to be exposed publicly.”

xAI did not respond to a request for comment. Nor did the 28-year-old xAI technical staff member whose key was exposed.

Carole Winqwist, chief marketing officer at GitGuardian, said giving potentially hostile users free access to private LLMs is a recipe for disaster.

“If you’re an attacker and you have direct access to the model and the back end interface for things like Grok, it’s definitely something you can use for further attacking,” she said. “An attacker could it use for prompt injection, to tweak the (LLM) model to serve their purposes, or try to implant code into the supply chain.”

The inadvertent exposure of internal LLMs for xAI comes as Musk’s so-called Department of Government Efficiency (DOGE) has been feeding sensitive government records into artificial intelligence tools. In February, The Washington Post reported DOGE officials were feeding data from across the Education Department into AI tools to probe the agency’s programs and spending.

The Post said DOGE plans to replicate this process across many departments and agencies, accessing the back-end software at different parts of the government and then using AI technology to extract and sift through information about spending on employees and programs.

“Feeding sensitive data into AI software puts it into the possession of a system’s operator, increasing the chances it will be leaked or swept up in cyberattacks,” Post reporters wrote.

Wired reported in March that DOGE has deployed a proprietary chatbot called GSAi to 1,500 federal workers at the General Services Administration, part of an effort to automate tasks previously done by humans as DOGE continues its purge of the federal workforce.

A Reuters report last month said Trump administration officials told some U.S. government employees that DOGE is using AI to surveil at least one federal agency’s communications for hostility to President Trump and his agenda. Reuters wrote that the DOGE team has heavily deployed Musk’s Grok AI chatbot as part of their work slashing the federal government, although Reuters said it could not establish exactly how Grok was being used.

Caturegli said while there is no indication that federal government or user data could be accessed through the exposed x.ai API key, these private models are likely trained on proprietary data and may unintentionally expose details related to internal development efforts at xAI, Twitter, or SpaceX.

“The fact that this key was publicly exposed for two months and granted access to internal models is concerning,” Caturegli said. “This kind of long-lived credential exposure highlights weak key management and insufficient internal monitoring, raising questions about safeguards around developer access and broader operational security.”

How Each Pillar of the 1st Amendment is Under Attack

“Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances.” -U.S. Constitution, First Amendment.

In an address to Congress this month, President Trump claimed he had “brought free speech back to America.” But barely two months into his second term, the president has waged an unprecedented attack on the First Amendment rights of journalists, students, universities, government workers, lawyers and judges.

This story explores a slew of recent actions by the Trump administration that threaten to undermine all five pillars of the First Amendment to the U.S. Constitution, which guarantees freedoms concerning speech, religion, the media, the right to assembly, and the right to petition the government and seek redress for wrongs.

THE RIGHT TO PETITION

The right to petition allows citizens to communicate with the government, whether to complain, request action, or share viewpoints — without fear of reprisal. But that right is being assaulted by this administration on multiple levels. For starters, many GOP lawmakers are now heeding their leadership’s advice to stay away from local town hall meetings and avoid the wrath of constituents affected by the administration’s many federal budget and workforce cuts.

Another example: President Trump recently fired most of the people involved in processing Freedom of Information Act (FOIA) requests for government agencies. FOIA is an indispensable tool used by journalists and the public to request government records, and to hold leaders accountable.

The biggest story by far this week was the bombshell from The Atlantic editor Jeffrey Goldberg, who recounted how he was inadvertently added to a Signal group chat with National Security Advisor Michael Waltz and 16 other Trump administration officials discussing plans for an upcoming attack on Yemen.

One overlooked aspect of Goldberg’s incredible account is that by planning and coordinating the attack on Signal — which features messages that can auto-delete after a short time — administration officials were evidently seeking a way to avoid creating a lasting (and potentially FOIA-able) record of their deliberations.

“Intentional or not, use of Signal in this context was an act of erasure—because without Jeffrey Goldberg being accidentally added to the list, the general public would never have any record of these communications or any way to know they even occurred,” Tony Bradley wrote this week at Forbes.

Petitioning the government, particularly when it ignores your requests, often requires challenging federal agencies in court. But that becomes far more difficult if the most competent law firms start to shy away from cases that may involve crossing the president and his administration.

On March 22, the president issued a memorandum that directs heads of the Justice and Homeland Security Departments to “seek sanctions against attorneys and law firms who engage in frivolous, unreasonable and vexatious litigation against the United States,” or in matters that come before federal agencies.

The POTUS recently issued several executive orders railing against specific law firms with attorneys who worked legal cases against him. On Friday, the president announced that the law firm of Skadden, Arps, Slate, Meager & Flom had agreed to provide $100 million in pro bono work on issues that he supports.

Trump issued another order naming the firm Paul, Weiss, Rifkind, Wharton & Garrison, which ultimately agreed to pledge $40 million in pro bono legal services to the president’s causes.

Other Trump executive orders targeted law firms Jenner & Block and WilmerHale, both of which have attorneys that worked with special counsel Robert Mueller on the investigation into Russian interference in the 2016 election. But this week, two federal judges in separate rulings froze parts of those orders.

“There is no doubt this retaliatory action chills speech and legal advocacy, and that is qualified as a constitutional harm,” wrote Judge Richard Leon, who ruled against the executive order targeting WilmerHale.

President Trump recently took the extraordinary step of calling for the impeachment of federal judges who rule against the administration. Trump called U.S. District Judge James Boasberg a “Radical Left Lunatic” and urged he be removed from office for blocking deportation of Venezuelan alleged gang members under a rarely invoked wartime legal authority.

In a rare public rebuke to a sitting president, U.S. Supreme Court Justice John Roberts issued a statement on March 18 pointing out that “For more than two centuries, it has been established that impeachment is not an appropriate response to disagreement concerning a judicial decision.”

The U.S. Constitution provides that judges can be removed from office only through impeachment by the House of Representatives and conviction by the Senate. The Constitution also states that judges’ salaries cannot be reduced while they are in office.

Undeterred, House Speaker Mike Johnson this week suggested the administration could still use the power of its purse to keep courts in line, and even floated the idea of wholesale eliminating federal courts.

“We do have authority over the federal courts as you know,” Johnson said. “We can eliminate an entire district court. We have power of funding over the courts, and all these other things. But desperate times call for desperate measures, and Congress is going to act, so stay tuned for that.”

FREEDOM OF ASSEMBLY

President Trump has taken a number of actions to discourage lawful demonstrations at universities and colleges across the country, threatening to cut federal funding for any college that supports protests he deems “illegal.”

A Trump executive order in January outlined a broad federal crackdown on what he called “the explosion of antisemitism” on U.S. college campuses. This administration has asserted that foreign students who are lawfully in the United States on visas do not enjoy the same free speech or due process rights as citizens.

Reuters reports that the acting civil rights director at the Department of Education on March 10 sent letters to 60 educational institutions warning they could lose federal funding if they don’t do more to combat anti-semitism. On March 20, Trump issued an order calling for the closure of the Education Department.

Meanwhile, U.S. Immigration and Customs Enforcement (ICE) agents have been detaining and trying to deport pro-Palestinian students who are legally in the United States. The administration is targeting students and academics who spoke out against Israel’s attacks on Gaza, or who were active in campus protests against U.S. support for the attacks. Secretary of State Marco Rubio told reporters Thursday that at least 300 foreign students have seen their visas revoked under President Trump, a far higher number than was previously known.

In his first term, Trump threatened to use the national guard or the U.S. military to deal with protesters, and in campaigning for re-election he promised to revisit the idea.

“I think the bigger problem is the enemy from within,” Trump told Fox News in October 2024. “We have some very bad people. We have some sick people, radical left lunatics. And I think they’re the big — and it should be very easily handled by, if necessary, by National Guard, or if really necessary, by the military, because they can’t let that happen.”

This term, Trump acted swiftly to remove the top judicial advocates in the armed forces who would almost certainly push back on any request by the president to use U.S. soldiers in an effort to quell public protests, or to arrest and detain immigrants. In late February, the president and Defense Secretary Pete Hegseth fired the top legal officers for the military services — those responsible for ensuring the Uniform Code of Military Justice is followed by commanders.

Military.com warns that the purge “sets an alarming precedent for a crucial job in the military, as President Donald Trump has mused about using the military in unorthodox and potentially illegal ways.” Hegseth told reporters the removals were necessary because he didn’t want them to pose any “roadblocks to orders that are given by a commander in chief.”

FREEDOM OF THE PRESS

President Trump has sued a number of U.S. news outlets, including 60 Minutes, CNN, The Washington Post, The New York Times and other smaller media organizations for unflattering coverage.

In a $10 billion lawsuit against 60 Minutes and its parent Paramount, Trump claims they selectively edited an interview with former Vice President Kamala Harris prior to the 2024 election. The TV news show last month published transcripts of the interview at the heart of the dispute, but Paramount is reportedly considering a settlement to avoid potentially damaging its chances of winning the administration’s approval for a pending multibillion-dollar merger.

The president sued The Des Moines Register and its parent company, Gannett, for publishing a poll showing Trump trailing Harris in the 2024 presidential election in Iowa (a state that went for Trump). The POTUS also is suing the Pulitzer Prize board over 2018 awards given to The New York Times and The Washington Post for their coverage of purported Russian interference in the 2016 election.

Whether or not any of the president’s lawsuits against news organizations have merit or succeed is almost beside the point. The strategy behind suing the media is to make reporters and newsrooms think twice about criticizing or challenging the president and his administration. The president also knows some media outlets will find it more expedient to settle.

Trump also sued ABC News and George Stephanopoulos for stating that the president had been found liable for “rape” in a civil case [Trump was found liable of sexually abusing and defaming E. Jean Carroll]. ABC parent Disney settled that claim by agreeing to donate $15 million to the Trump Presidential Library.

Following the attack on the U.S. Capitol on Jan. 6, 2021, Facebook blocked President Trump’s account. Trump sued Meta, and after the president’s victory in 2024 Meta settled and agreed to pay Trump $25 million: $22 million would go to his presidential library, and the rest to legal fees. Meta CEO Mark Zuckerberg also announced Facebook and Instagram would get rid of fact-checkers and rely instead on reader-submitted “community notes” to debunk disinformation on the social media platform.

Brendan Carr, the president’s pick to run the Federal Communications Commission (FCC), has pledged to “dismantle the censorship cartel and restore free speech rights for everyday Americans.” But on January 22, 2025, the FCC reopened complaints against ABC, CBS and NBC over their coverage of the 2024 election. The previous FCC chair had dismissed the complaints as attacks on the First Amendment and an attempt to weaponize the agency for political purposes.

According to Reuters, the complaints call for an investigation into how ABC News moderated the pre-election TV debate between Trump and Biden, and appearances of then-Vice President Harris on 60 Minutes and on NBC’s “Saturday Night Live.”

Since then, the FCC has opened investigations into NPR and PBS, alleging that they are breaking sponsorship rules. The Center for Democracy & Technology (CDT), a think tank based in Washington, D.C., noted that the FCC is also investigating KCBS in San Francisco for reporting on the location of federal immigration authorities.

“Even if these investigations are ultimately closed without action, the mere fact of opening them – and the implicit threat to the news stations’ license to operate – can have the effect of deterring the press from news coverage that the Administration dislikes,” the CDT’s Kate Ruane observed.

Trump has repeatedly threatened to “open up” libel laws, with the goal of making it easier to sue media organizations for unfavorable coverage. But this week, the U.S. Supreme Court declined to hear a challenge brought by Trump donor and Las Vegas casino magnate Steve Wynn to overturn the landmark 1964 decision in New York Times v. Sullivan, which insulates the press from libel suits over good-faith criticism of public figures.

The president also has insisted on picking which reporters and news outlets should be allowed to cover White House events and participate in the press pool that trails the president. He barred the Associated Press from the White House and Air Force One over their refusal to call the Gulf of Mexico by another name.

And the Defense Department has ordered a number of top media outlets to vacate their spots at the Pentagon, including CNN, The Hill, The Washington Post, The New York Times, NBC News, Politico and National Public Radio.

“Incoming media outlets include the New York Post, Breitbart, the Washington Examiner, the Free Press, the Daily Caller, Newsmax, the Huffington Post and One America News Network, most of whom are seen as conservative or favoring Republican President Donald Trump,” Reuters reported.

FREEDOM OF SPEECH

Shortly after Trump took office again in January 2025, the administration began circulating lists of hundreds of words that government staff and agencies shall not use in their reports and communications.

The Brookings Institution notes that in moving to comply with this anti-speech directive, federal agencies have purged countless taxpayer-funded data sets from a swathe of government websites, including data on crime, sexual orientation, gender, education, climate, and global development.

The New York Times reports that in the past two months, hundreds of terabytes of digital resources analyzing data have been taken off government websites.

“While in many cases the underlying data still exists, the tools that make it possible for the public and researchers to use that data have been removed,” The Times wrote.

On Jan. 27, Trump issued a memo (PDF) that paused all federally funded programs pending a review of those programs for alignment with the administration’s priorities. Among those was ensuring that no funding goes toward advancing “Marxist equity, transgenderism, and green new deal social engineering policies.”

According to the CDT, this order is a blatant attempt to force government grantees to cease engaging in speech that the current administration dislikes, including speech about the benefits of diversity, climate change, and LGBTQ issues.

“The First Amendment does not permit the government to discriminate against grantees because it does not like some of the viewpoints they espouse,” the CDT’s Ruane wrote. “Indeed, those groups that are challenging the constitutionality of the order argued as much in their complaint, and have won an injunction blocking its implementation.”

On January 20, the same day Trump issued an executive order on free speech, the president also issued an executive order titled “Reevaluating and Realigning United States Foreign Aid,” which froze funding for programs run by the U.S. Agency for International Development (USAID). Among those were programs designed to empower civil society and human rights groups, journalists and others responding to digital repression and Internet shutdowns.

According to the Electronic Frontier Foundation (EFF), this includes many freedom technologies that use cryptography, fight censorship, protect freedom of speech, privacy and anonymity for millions of people around the world.

“While the State Department has issued some limited waivers, so far those waivers do not seem to cover the open source internet freedom technologies,” the EFF wrote about the USAID disruptions. “As a result, many of these projects have to stop or severely curtail their work, lay off talented workers, and stop or slow further development.”

On March 14, the president signed another executive order that effectively gutted the U.S. Agency for Global Media (USAGM), which oversees or funds media outlets including Radio Free Europe/Radio Liberty and Voice of America (VOA). The USAGM also oversees Radio Free Asia, which supporters say has been one of the most reliable tools used by the government to combat Chinese propaganda.

But this week, U.S. District Court Judge Royce Lamberth, a Reagan appointee, temporarily blocked USAGM’s closure by the administration.

“RFE/RL has, for decades, operated as one of the organizations that Congress has statutorily designated to carry out this policy,” Lamberth wrote in a 10-page opinion. “The leadership of USAGM cannot, with one sentence of reasoning offering virtually no explanation, force RFE/RL to shut down — even if the President has told them to do so.”

FREEDOM OF RELIGION

The Trump administration rescinded a decades-old policy that instructed officers not to take immigration enforcement actions in or near “sensitive” or “protected” places, such as churches, schools, and hospitals.

That directive was immediately challenged in a case brought by a group of Quakers, Baptists and Sikhs, who argued the policy reversal was keeping people from attending services for fear of being arrested on civil immigration violations. On Feb. 24, a federal judge agreed and blocked ICE agents from entering churches or targeting migrants nearby.

The president’s executive order allegedly addressing antisemitism came with a fact sheet that described college campuses as “infested” with “terrorists” and “jihadists.” Multiple faith groups expressed alarm over the order, saying it attempts to weaponize antisemitism and promote “dehumanizing anti-immigrant policies.”

The president also announced the creation of a “Task Force to Eradicate Anti-Christian Bias,” to be led by Attorney General Pam Bondi. Never mind that Christianity is easily the largest faith in America and that Christians are well-represented in Congress.

The Rev. Paul Brandeis Raushenbush, a Baptist minister and head of the progressive Interfaith Alliance, issued a statement accusing Trump of hypocrisy in claiming to champion religion by creating the task force.

“From allowing immigration raids in churches, to targeting faith-based charities, to suppressing religious diversity, the Trump Administration’s aggressive government overreach is infringing on religious freedom in a way we haven’t seen for generations,” Raushenbush said.

A statement from Americans United for Separation of Church and State said the task force could lead to religious persecution of those with other faiths.

“Rather than protecting religious beliefs, this task force will misuse religious freedom to justify bigotry, discrimination, and the subversion of our civil rights laws,” said Rachel Laser, the group’s president and CEO.

Where is President Trump going with all these blatant attacks on the First Amendment? The president has made no secret of his affection for autocratic leaders and “strongmen” around the world, and he is particularly enamored with Hungary’s far-right Prime Minister Viktor Orbán, who has visited Trump’s Mar-a-Lago resort twice in the past year.

A March 15 essay in The Atlantic by Hungarian investigative journalist András Pethő recounts how Orbán rose to power by consolidating control over the courts, and by building his own media universe while simultaneously placing a stranglehold on the independent press.

“As I watch from afar what’s happening to the free press in the United States during the first weeks of Trump’s second presidency — the verbal bullying, the legal harassment, the buckling by media owners in the face of threats — it all looks very familiar,” Pethő wrote. “The MAGA authorities have learned Orbán’s lessons well.”

DOGE to Fired CISA Staff: Email Us Your Personal Data

A message posted on Monday to the homepage of the U.S. Cybersecurity & Infrastructure Security Agency (CISA) is the latest exhibit in the Trump administration’s continued disregard for basic cybersecurity protections. The message instructed recently-fired CISA employees to get in touch so they can be rehired and then immediately placed on leave, asking employees to send their Social Security number or date of birth in a password-protected email attachment — presumably with the password needed to view the file included in the body of the email.

On March 13, a Maryland district court judge ordered the Trump administration to reinstate more than 130 probationary CISA employees who were fired last month. On Monday, the administration announced that those dismissed employees would be reinstated but placed on paid administrative leave. They are among nearly 25,000 fired federal workers who are in the process of being rehired.

A notice covering the CISA homepage said the administration is making every effort to contact those who were unlawfully fired in mid-February.

“Please provide a password protected attachment that provides your full name, your dates of employment (including date of termination), and one other identifying factor such as date of birth or social security number,” the message reads. “Please, to the extent that it is available, attach any termination notice.”

The message didn’t specify how affected CISA employees should share the password for any attached files, so the implicit expectation is that employees should just include the plaintext password in their message.

Email is about as secure as a postcard sent through the mail, because anyone who manages to intercept the missive anywhere along its path of delivery can likely read it. In security terms, that’s the equivalent of encrypting sensitive data while also attaching the secret key needed to view the information.

What’s more, a great many antivirus and security scanners have trouble inspecting password-protected files, meaning the administration’s instructions are likely to increase the risk that malware submitted by cybercriminals could be accepted and opened by U.S. government employees.

The message in the screenshot above was removed from the CISA homepage Tuesday evening and replaced with a much shorter notice directing former CISA employees to contact a specific email address. But a slightly different version of the same message originally posted to CISA’s website still exists at the website for the U.S. Citizenship and Immigration Services, which likewise instructs those fired employees who wish to be rehired and put on leave to send a password-protected email attachment with sensitive personal data.

This is hardly the first example of the administration discarding Security 101 practices in the name of expediency. Last month, the Central Intelligence Agency (CIA) sent an unencrypted email to the White House with the first names and first letter of the last names of recently hired CIA officers who might be easy to fire.

As cybersecurity journalist Shane Harris noted in The Atlantic, even those fragments of information could be useful to foreign spies.

“Over the weekend, a former senior CIA official showed me the steps by which a foreign adversary who knew only his first name and last initial could have managed to identify him from the single line of the congressional record where his full name was published more than 20 years ago, when he became a member of the Foreign Service,” Harris wrote. “The former official was undercover at the time as a State Department employee. If a foreign government had known even part of his name from a list of confirmed CIA officers, his cover would have been blown.”

The White House has also fired at least 100 intelligence staffers from the National Security Agency (NSA), reportedly for using an internal NSA chat tool to discuss their personal lives and politics. Testifying before the House Select Committee on the Communist Party earlier this month, the NSA’s former top cybersecurity official said the Trump administration’s attempts to mass fire probationary federal employees will be “devastating” to U.S. cybersecurity operations.

Rob Joyce, who spent 34 years at the NSA, told Congress how important those employees are in sustaining an aggressive stance against China in cyberspace.

“At my former agency, remarkable technical talent was recruited into developmental programs that provided intensive unique training and hands-on experience to cultivate vital skills,” Joyce told the panel. “Eliminating probationary employees will destroy a pipeline of top talent responsible for hunting and eradicating [Chinese] threats.”

Both the message to fired CISA workers and DOGE’s ongoing efforts to bypass vetted government networks for a faster Wi-Fi signal are emblematic of this administration’s overall approach to even basic security measures: To go around them, or just pretend they don’t exist for a good reason.

On Monday, The New York Times reported that U.S. Secret Service agents at the White House were briefly on alert last month when a trusted captain of Elon Musk’s “Department of Government Efficiency” (DOGE) visited the roof of the Eisenhower building inside the White House compound — to see about setting up a dish to receive satellite Internet access directly from Musk’s Starlink service.

The White House press secretary told The Times that Starlink had “donated” the service and that the gift had been vetted by the lawyer overseeing ethics issues in the White House Counsel’s Office. The White House claims the service is necessary because its wireless network is too slow.

Jake Williams, vice president for research and development at the cybersecurity consulting firm Hunter Strategy, told The Times “it’s super rare” to install Starlink or another internet provider as a replacement for existing government infrastructure that has been vetted and secured.

“I can’t think of a time that I have heard of that,” Williams said. “It introduces another attack point,” Williams said. “But why introduce that risk?”

Meanwhile, NBC News reported on March 7 that Starlink is expanding its footprint across the federal government.

“Multiple federal agencies are exploring the idea of adopting SpaceX’s Starlink for internet access — and at least one agency, the General Services Administration (GSA), has done so at the request of Musk’s staff, according to someone who worked at the GSA last month and is familiar with its network operations — despite a vow by Musk and Trump to slash the overall federal budget,” NBC wrote.

The longtime Musk employee who encountered the Secret Service on the roof in the White House complex was Christopher Stanley, the 33-year-old senior director for security engineering at X and principal security engineer at SpaceX.

On Monday, Bloomberg broke the news that Stanley had been tapped for a seat on the board of directors at the mortgage giant Fannie Mae. Stanley was added to the board alongside newly confirmed Federal Housing Finance Agency director Bill Pulte, the grandson of the late housing businessman and founder of PulteGroup — William J. Pulte.

In a nod to his new board role atop an agency that helps drive the nation’s $12 trillion mortgage market, Stanley retweeted a Bloomberg story about the hire with a smiley emoji and the comment “Tech Support.”

But earlier today, Bloomberg reported that Stanley had abruptly resigned from the Fannie board, and that details about the reason for his quick departure weren’t immediately clear. As first reported here last month, Stanley had a brush with celebrity on Twitter in 2015 when he leaked the user database for the DDoS-for-hire service LizardStresser, and soon faced threats of physical violence against his family.

My 2015 story on that leak did not name Stanley, but he exposed himself as the source by posting a video about it on his Youtube channel. A review of domain names registered by Stanley shows he went by the nickname “enKrypt,” and was the former owner of a pirated software and hacking forum called error33[.]net, as well as theC0re, a video game cheating community.

Stanley is one of more than 50 DOGE workers, mostly young men and women who have worked with one or more of Musk’s companies. The Trump administration remains dogged by questions about how many — if any — of the DOGE workers were put through the gauntlet of a thorough security background investigation before being given access to such sensitive government databases.

That’s largely because in one of his first executive actions after being sworn in for a second term on Jan. 20, President Trump declared that the security clearance process was simply too onerous and time-consuming, and that anyone so designated by the White House counsel would have full top secret/sensitive compartmented information (TS/SCI) clearances for up to six months. Translation: We accepted the risk, so TAH-DAH! No risk!

Presumably, this is the same counsel who saw no ethical concerns with Musk “donating” Starlink to the White House, or with President Trump summoning the media to film him hawking Cybertrucks and Teslas (a.k.a. “Teslers”) on the White House lawn last week.

Mr. Musk’s unelected role as head of an ad hoc executive entity that is gleefully firing federal workers and feeding federal agencies into “the wood chipper” has seen his Tesla stock price plunge in recent weeks, while firebombings and other vandalism attacks on property carrying the Tesla logo are cropping up across the U.S. and overseas and driving down Tesla sales.

President Trump and his attorney general Pam Bondi have dubiously asserted that those responsible for attacks on Tesla dealerships are committing “domestic terrorism,” and that vandals will be prosecuted accordingly. But it’s not clear this administration would recognize a real domestic security threat if it was ensconced squarely behind the Resolute Desk.

Or at the pinnacle of the Federal Bureau of Investigation (FBI). The Washington Post reported last month that Trump’s new FBI director Kash Patel was paid $25,000 last year by a film company owned by a dual U.S. Russian citizen that has made programs promoting “deep state” conspiracy theories pushed by the Kremlin.

“The resulting six-part documentary appeared on Tucker Carlson’s online network, itself a reliable conduit for Kremlin propaganda,” The Post reported. “In the film, Patel made his now infamous pledge to shut down the FBI’s headquarters in Washington and ‘open it up as a museum to the deep state.'”

When the head of the FBI is promising to turn his own agency headquarters into a mocking public exhibit on the U.S. National Mall, it may seem silly to fuss over the White House’s clumsy and insulting instructions to former employees they unlawfully fired.

Indeed, one consistent feedback I’ve heard from a subset of readers here is something to this effect: “I used to like reading your stuff more when you weren’t writing about politics all the time.”

My response to that is: “Yeah, me too.” It’s not that I’m suddenly interested in writing about political matters; it’s that various actions by this administration keep intruding on my areas of coverage.

A less charitable interpretation of that reader comment is that anyone still giving such feedback is either dangerously uninformed, being disingenuous, or just doesn’t want to keep being reminded that they’re on the side of the villains, despite all the evidence showing it.

Article II of the U.S. Constitution unambiguously states that the president shall take care that the laws be faithfully executed. But almost from Day One of his second term, Mr. Trump has been acting in violation of his sworn duty as president by choosing not to enforce laws passed by Congress (TikTok ban, anyone?), by freezing funds already allocated by Congress, and most recently by flouting a federal court order while simultaneously calling for the impeachment of the judge who issued it. Sworn to uphold, protect and defend The Constitution, President Trump appears to be creating new constitutional challenges with almost each passing day.

When Mr. Trump was voted out of office in November 2020, he turned to baseless claims of widespread “election fraud” to explain his loss — with deadly and long-lasting consequences. This time around, the rallying cry of DOGE and White House is “government fraud,” which gives the administration a certain amount of cover for its actions among a base of voters that has long sought to shrink the size and cost of government.

In reality, “government fraud” has become a term of derision and public scorn applied to anything or anyone the current administration doesn’t like. If DOGE and the White House were truly interested in trimming government waste, fraud and abuse, they could scarcely do better than consult the inspectors general fighting it at various federal agencies.

After all, the inspectors general likely know exactly where a great deal of the federal government’s fiscal skeletons are buried. Instead, Mr. Trump fired at least 17 inspectors general, leaving the government without critical oversight of agency activities. That action is unlikely to stem government fraud; if anything, it will only encourage such activity.

As Techdirt founder Mike Masnick noted in a recent column “Why Techdirt is Now a Democracy Blog (Whether We Like it or Not),” when the very institutions that made American innovation possible are being systematically dismantled, it’s not a “political” story anymore: It’s a story about whether the environment that enabled all the other stories we cover will continue to exist.

“This is why tech journalism’s perspective is so crucial right now,” Masnick wrote. “We’ve spent decades documenting how technology and entrepreneurship can either strengthen or undermine democratic institutions. We understand the dangers of concentrated power in the digital age. And we’ve watched in real-time as tech leaders who once championed innovation and openness now actively work to consolidate control and dismantle the very systems that enabled their success.”

“But right now, the story that matters most is how the dismantling of American institutions threatens everything else we cover,” Masnick continued. “When the fundamental structures that enable innovation, protect civil liberties, and foster open dialogue are under attack, every other tech policy story becomes secondary.”

Trump 2.0 Brings Cuts to Cyber, Consumer Protections

One month into his second term, President Trump’s actions to shrink the government through mass layoffs, firings and withholding funds allocated by Congress have thrown federal cybersecurity and consumer protection programs into disarray. At the same time, agencies are battling an ongoing effort by the world’s richest man to wrest control over their networks and data.

The Trump administration has fired at least 130 employees at the federal government’s foremost cybersecurity body — the Cybersecurity and Infrastructure Security Agency (CISA). Those dismissals reportedly included CISA staff dedicated to securing U.S. elections, and fighting misinformation and foreign influence operations.

Earlier this week, technologists with Elon Musk’s Department of Government Efficiency (DOGE) arrived at CISA and gained access to the agency’s email and networked files. Those DOGE staffers include Edward “Big Balls” Coristine, a 19-year-old former denizen of the “Com,” an archipelago of Discord and Telegram chat channels that function as a kind of distributed cybercriminal social network.

The investigative journalist Jacob Silverman writes that Coristine is the grandson of Valery Martynov, a KGB double agent who spied for the United States. Silverman recounted how Martynov’s wife Natalya Martynova moved to the United States with her two children after her husband’s death.

“Her son became a Virginia police officer who sometimes posts comments on blogs about his historically famous father,” Silverman wrote. “Her daughter became a financial professional who married Charles Coristine, the proprietor of LesserEvil, a snack company. Among their children is a 19-year-old young man named Edward Coristine, who currently wields an unknown amount of power and authority over the inner-workings of our federal government.”

Another member of DOGE is Christopher Stanley, formerly senior director for security engineering at X and principal security engineer at Musk’s SpaceX. Stanley, 33, had a brush with celebrity on Twitter in 2015 when he leaked the user database for the DDoS-for-hire service LizardStresser, and soon faced threats of physical violence against his family.

My 2015 story on that leak did not name Stanley, but he exposed himself as the source by posting a video about it on his Youtube channel. A review of domain names registered by Stanley shows he went by the nickname “enKrypt,” and was the former owner of a pirated software and hacking forum called error33[.]net, as well as theC0re, a video game cheating community.

“A NATIONAL CYBERATTACK”

DOGE has been steadily gaining sensitive network access to federal agencies that hold a staggering amount of personal and financial information on Americans, including the Social Security Administration (SSA), the Department of Homeland Security, the Office of Personnel Management (OPM), and the Treasury Department.

Most recently, DOGE has sought broad access to systems at the Internal Revenue Service that contain the personal tax information on millions of Americans, including how much individuals earn and owe, property information, and even details related to child custody agreements. The New York Times reported Friday that the IRS had reached an agreement whereby a single DOGE employee — 25-year-old Gavin Kliger — will be allowed to see only anonymized taxpayer information.

The rapidity with which DOGE has rifled through one federal database after another in the name of unearthing “massive fraud” by government agencies has alarmed many security experts, who warned that DOGE’s actions bypassed essential safeguards and security measures.

“The most alarming aspect isn’t just the access being granted,” wrote Bruce Schneier and Davi Ottenheimer, referring to DOGE as a national cyberattack. “It’s the systematic dismantling of security measures that would detect and prevent misuse—including standard incident response protocols, auditing, and change-tracking mechanisms—by removing the career officials in charge of those security measures and replacing them with inexperienced operators.”

Jacob Williams is a former hacker with the U.S. National Security Agency who now works as managing director of the cybersecurity firm Hunter Labs. Williams kicked a virtual hornet’s nest last week when he posted on LinkedIn that the network incursions by DOGE were “a bigger threat to U.S. federal government information systems than China.”

Williams said while he doesn’t believe anyone at DOGE would intentionally harm the integrity and availability of these systems, it’s widely reported (and not denied) that DOGE introduced code changes into multiple federal IT systems. These code changes, he maintained, are not following the normal process for vetting and review given to federal government IT systems.

“For those thinking ‘I’m glad they aren’t following the normal federal government IT processes, those are too burdensome’ I get where you’re coming from,” Williams wrote. “But another name for ‘red tape’ are ‘controls.’ If you’re comfortable bypassing controls for the advancement of your agenda, I have questions – mostly about whether you do this in your day job too. Please tag your employer letting them know your position when you comment that controls aren’t important (doubly so if you work in cybersecurity). All satire aside, if you’re comfortable abandoning controls for expediency, I implore you to decide where the line is that you won’t cross in that regard.”

The DOGE website’s “wall of receipts” boasts that Musk and his team have saved the federal government more than $55 billion through staff reductions, lease cancellations and terminated contracts. But a team of reporters at The New York Times found the math that could back up those checks is marred with accounting errors, incorrect assumptions, outdated data and other mistakes.

For example, DOGE claimed it saved $8 billion in one contract, when the total amount was actually $8 million, The Times found.

“Some contracts the group claims credit for were double- or triple-counted,” reads a Times story with six bylines. “Another initially contained an error that inflated the totals by billions of dollars. While the DOGE team has surely cut some number of billions of dollars, its slapdash accounting adds to a pattern of recklessness by the group, which has recently gained access to sensitive government payment systems.”

So far, the DOGE website does not inspire confidence: We learned last week that the doge.gov administrators somehow left their database wide open, allowing someone to publish messages that ridiculed the site’s insecurity.

APPOINTMENTS

Trump’s efforts to grab federal agencies by their data has seen him replace career civil servants who refused to allow DOGE access to agency networks. CNN reports that Michelle King, acting commissioner of the Social Security Administration for more than 30 years, was shown the door after she denied DOGE access to sensitive information.

King was replaced by Leland Dudek, formerly a senior advisor in the SSA’s Office of Program Integrity. This week, Dudek posted a now-deleted message on LinkedIn acknowledging he had been placed on administrative leave for cooperating with DOGE.

“I confess,” Dudek wrote. “I bullied agency executives, shared executive contact information, and circumvented the chain of command to connect DOGE with the people who get stuff done. I confess. I asked where the fat was and is in our contracts so we can make the right tough choices.”

According to Wired, the National Institute of Standards and Technology (NIST) was also bracing this week for roughly 500 staffers to be fired, which could have serious impacts on NIST’s cybersecurity standards and software vulnerability tracking work.

“And cuts last week at the US Digital Service included the cybersecurity lead for the central Veterans Affairs portal, VA.gov, potentially leaving VA systems and data more vulnerable without someone in his role,” Wired’s Andy Greenberg and Lily Hay Newman wrote.

NextGov reports that Trump named the Department of Defense’s new chief information security officer: Katie Arrington, a former South Carolina state lawmaker who helped steer Pentagon cybersecurity contracting policy before being put on leave amid accusations that she disclosed classified data from a military intelligence agency.

NextGov notes that the National Security Agency suspended her clearance in 2021, although the exact reasons that led to the suspension and her subsequent leave were classified. Arrington argued that the suspension was a politically motivated effort to silence her.

Trump also appointed the former chief operating officer of the Republican National Committee as the new head of the Office of National Cyber Director. Sean Cairncross, who has no formal experience in technology or security, will be responsible for coordinating national cybersecurity policy, advising the president on cyber threats, and ensuring a unified federal response to emerging cyber-risks, Politico writes.

DarkReading reports that Cairncross would share responsibility for advising the president on cyber matters, along with the director of cyber at the White House National Security Council (NSC) — a group that advises the president on all matters security related, and not just cyber.

CONSUMER PROTECTION?

The president also ordered staffers at the Consumer Financial Protection Bureau (CFPB) to stop most work. Created by Congress in 2011 to be a clearinghouse of consumer complaints, the CFPB has sued some of the nation’s largest financial institutions for violating consumer protection laws.

The CFPB says its actions have put nearly $18 billion back in Americans’ pockets in the form of monetary compensation or canceled debts, and imposed $4 billion in civil money penalties against violators. The CFPB’s homepage has featured a “404: Page not found” error for weeks now.

Trump has appointed Russell Vought, the architect of the conservative policy playbook Project 2025, to be the CFPB’s acting director. Vought has publicly favored abolishing the agency, as has Elon Musk, whose efforts to remake X into a payments platform would otherwise be regulated by the CFPB.

The New York Times recently published a useful graphic showing all of the government staffing changes, including the firing of several top officials, affecting agencies with federal investigations into or regulatory battles with Musk’s companies. Democrats on the House Judiciary Committee also have released a comprehensive account (PDF) of Musk’s various conflicts of interest.

As the Times notes, Musk and his companies have repeatedly failed to comply with federal reporting protocols aimed at protecting state secrets, and these failures have prompted at least three federal reviews. Those include an inquiry launched last year by the Defense Department’s Office of Inspector General. Four days after taking office, Trump fired the DoD inspector general along with 17 other inspectors general.

The Trump administration also shifted the enforcement priorities of the U.S. Securities and Exchange Commission (SEC) away from prosecuting misconduct in the cryptocurrency sector, reassigning lawyers and renaming the unit to focus more on “cyber and emerging technologies.”

Reuters reports that the former SEC chair Gary Gensler made fighting misconduct in a sector he termed the “wild west” a priority for the agency, targeting not only cryptocurrency fraudsters but also the large firms that facilitate trading such as Coinbase.

On Friday, Coinbase said the SEC planned to withdraw its lawsuit against the crypto exchange. Also on Friday, the cryptocurrency exchange Bybit announced on X that a cybersecurity breach led to the theft of more than $1.4 billion worth of cryptocurrencies — making it the largest crypto heist ever.

ORGANIZED CRIME AND CORRUPTION

On Feb. 10, Trump ordered executive branch agencies to stop enforcing the U.S. Foreign Corrupt Practices Act, which froze foreign bribery investigations, and even allows for “remedial actions” of past enforcement actions deemed “inappropriate.”

Trump’s action also disbanded the Kleptocracy Asset Recovery Initiative and KleptoCapture Task Force — units which proved their value in corruption cases and in seizing the assets of sanctioned Russian oligarchs — and diverted resources away from investigating white-collar crime.

That’s according to the independent Organized Crime and Corruption Reporting Project (OCCRP), an investigative journalism outlet that until very recently was funded in part by the U.S. Agency for International Development (USAID).

The OCCRP lost nearly a third of its funding and was forced to lay off 43 reporters and staff after Trump moved to shutter USAID and freeze its spending. NBC News reports the Trump administration plans to gut the agency and leave fewer than 300 staffers on the job out of the current 8,000 direct hires and contractors.

The Global Investigative Journalism Network wrote this week that the sudden hold on USAID foreign assistance funding has frozen an estimated $268 million in agreed grants for independent media and the free flow of information in more than 30 countries — including several under repressive regimes.

Elon Musk has called USAID “a criminal organization” without evidence, and promoted fringe theories on his social media platform X that the agency operated without oversight and was rife with fraud. Just months before the election, USAID’s Office of Inspector General announced an investigation into USAID’s oversight of Starlink satellite terminals provided to the government of Ukraine.

KrebsOnSecurity this week heard from a trusted source that all outgoing email from USAID now carries a notation of “sensitive but unclassified,” a designation that experts say could make it more difficult for journalists and others to obtain USAID email records under the Freedom of Information Act (FOIA). On Feb. 20, Fedscoop reported also hearing the same thing from multiple sources, noting that the added message cannot be seen by senders until after the email is sent.

FIVE BULLETS

On Feb. 18, Trump issued an executive order declaring that only the U.S. attorney general and the president can provide authoritative interpretations of the law for the executive branch, and that this authority extends to independent agencies operating under the executive branch.

Trump is arguing that Article II, Clause 1 of the Constitution vests this power with the president. However, jurist.org writes that Article II does not expressly state the president or any other person in the executive branch has the power to interpret laws.

“The article states that the president is required to ‘take care that the laws be faithfully executed,'” Juris noted. “Jurisdiction to interpret laws and determine constitutionality belongs to the judicial branch under Article III. The framers of the Constitution designed the separation of duties to prevent any single branch of government from becoming too powerful.”

The executive order requires all agencies to submit to “performance standards and management objectives” to be established by the White House Office of Management and Budget, and to report periodically to the president.

Those performance metrics are already being requested: Employees at multiple federal agencies on Saturday reported receiving an email from the Office of Personnel Management ordering them to reply with a set of bullet points justifying their work for the past week.

“Please reply to this email with approx. 5 bullets of what you accomplished last week and cc your manager,” the notice read. “Please do not send any classified information, links, or attachments. Deadline is this Monday at 11:59 p.m. EST.”

In a social media post Saturday, Musk said the directive came at the behest of President Trump, and that failure to respond would be taken as a resignation. Meanwhile, Bloomberg writes the Department of Justice has been urging employees to hold off replying out of concern doing so could trigger ethics violations. The National Treasury Employees Union also is advising its employees not to respond.

A legal battle over Trump’s latest executive order is bound to join more than 70 other lawsuits currently underway to halt the administration’s efforts to massively reduce the size of the federal workforce through layoffs, firings and attrition.

KING TRUMP?

On Feb. 15, the president posted on social media, “He who saves his Country does not violate any Law,” citing a quote often attributed to the French dictator Napoleon Bonaparte. Four days later, Trump referred to himself as “the king” on social media, while the White House nonchalantly posted an illustration of him wearing a crown.

Trump has been publicly musing about running for an unconstitutional third-term in office, a statement that some of his supporters dismiss as Trump just trying to rile his liberal critics. However, just days after Trump began his second term, Rep. Andy Ogles (R-Tenn.) introduced a bill to amend the Constitution so that Trump — and any other future president — can be elected to serve a third term.

This week at the Conservative Political Action Conference (CPAC), Rep. Ogles reportedly led a group of Trump supporters calling itself the “Third Term Project,” which is trying to gain support for the bill from GOP lawmakers. The event featured images of Trump depicted as Caesar.

Russia continues to be among the world’s top exporters of cybercrime, narcotics, money laundering, human trafficking, disinformation, war and death, and yet the Trump administration has suddenly broken with the Western world in normalizing relations with Moscow.

This week President Trump stunned U.S. allies by repeating Kremlin talking points that Ukraine is somehow responsible for Russia’s invasion, and that Ukrainian President Volodymyr Zelensky is a “dictator.” The president repeated these lies even as his administration is demanding that Zelensky give the United States half of his country’s mineral wealth in exchange for a promise that Russia will cease its territorial aggression there.

President Trump’s servility toward an actual dictator — Russian President Vladimir Putin — does not bode well for efforts to improve the cybersecurity of U.S. federal IT networks, or the private sector systems on which the government is largely reliant. In addition, this administration’s baffling moves to alienate, antagonize and sideline our closest allies could make it more difficult for the United States to secure their ongoing cooperation in cybercrime investigations.

It’s also startling how closely DOGE’s approach so far hews to tactics typically employed by ransomware gangs: A group of 20-somethings with names like “Big Balls” shows up on a weekend and gains access to your servers, deletes data, locks out key staff, takes your website down, and prevents you from serving customers.

When the federal executive starts imitating ransomware playbooks against its own agencies while Congress largely gazes on in either bewilderment or amusement, we’re in four-alarm fire territory. At least in theory, one can negotiate with ransomware purveyors.

Who’s Behind the Seized Forums ‘Cracked’ & ‘Nulled’?

The FBI joined authorities across Europe last week in seizing domain names for Cracked and Nulled, English-language cybercrime forums with millions of users that trafficked in stolen data, hacking tools and malware. An investigation into the history of these communities shows their apparent co-founders quite openly operate an Internet service provider and a pair of e-commerce platforms catering to buyers and sellers on both forums.

On Jan. 30, the U.S. Department of Justice said it seized eight domain names that were used to operate Cracked, a cybercrime forum that sprang up in 2018 and attracted more than four million users. The DOJ said the law enforcement action, dubbed Operation Talent, also seized domains tied to Sellix, Cracked’s payment processor.

In addition, the government seized the domain names for two popular anonymity services that were heavily advertised on Cracked and Nulled and allowed customers to rent virtual servers: StarkRDP[.]io, and rdp[.]sh.

Those archived webpages show both RDP services were owned by an entity called 1337 Services Gmbh. According to corporate records compiled by Northdata.com, 1337 Services GmbH is also known as AS210558 and is incorporated in Hamburg, Germany.

The Cracked forum administrator went by the nicknames “FlorainN” and “StarkRDP” on multiple cybercrime forums. Meanwhile, a LinkedIn profile for a Florian M. from Germany refers to this person as the co-founder of Sellix and founder of 1337 Services GmbH.

Northdata’s business profile for 1337 Services GmbH shows the company is controlled by two individuals: 32-year-old Florian Marzahl and Finn Alexander Grimpe, 28.

Neither Marzahl nor Grimpe responded to requests for comment. But Grimpe’s first name is interesting because it corresponds to the nickname chosen by the founder of Nulled, who goes by the monikers “Finn” and “Finndev.” NorthData reveals that Grimpe was the founder of a German entity called DreamDrive GmbH, which rented out high-end sports cars and motorcycles.

According to the cyber intelligence firm Intel 471, a user named Finndev registered on multiple cybercrime forums, including Raidforums [seized by the FBI in 2022], Void[.]to, and vDOS, a DDoS-for-hire service that was shut down in 2016 after its founders were arrested.

The email address used for those accounts was f.grimpe@gmail.com. DomainTools.com reports f.grimpe@gmail.com was used to register at least nine domain names, including nulled[.]lol and nulled[.]it. Neither of these domains were among those seized in Operation Talent.

Intel471 finds the user FlorainN registered across multiple cybercrime forums using the email address olivia.messla@outlook.de. The breach tracking service Constella Intelligence says this email address used the same password (and slight variations of it) across many accounts online — including at hacker forums — and that the same password was used in connection with dozens of other email addresses, such as florianmarzahl@hotmail.de, and fmarzahl137@gmail.com.

The Justice Department said the Nulled marketplace had more than five million members, and has been selling stolen login credentials, stolen identification documents and hacking services, as well as tools for carrying out cybercrime and fraud, since 2016.

Perhaps fittingly, both Cracked and Nulled have been hacked over the years, exposing countless private messages between forum users. A review of those messages archived by Intel 471 showed that dozens of early forum members referred privately to Finndev as the owner of shoppy[.]gg, an e-commerce platform that caters to the same clientele as Sellix.

Shoppy was not targeted as part of Operation Talent, and its website remains online. Northdata reports that Shoppy’s business name — Shoppy Ecommerce Ltd. — is registered at an address in Gan-Ner, Israel, but there is no ownership information about this entity. Shoppy did not respond to requests for comment.

Constella found that a user named Shoppy registered on Cracked in 2019 using the email address finn@shoppy[.]gg. Constella says that email address is tied to a Twitter/X account for Shoppy Ecommerce in Israel.

The DOJ said one of the alleged administrators of Nulled, a 29-year-old Argentinian national named Lucas Sohn, was arrested in Spain. The government has not announced any other arrests or charges associated with Operation Talent.

Indeed, both StarkRDP and FloraiN have posted to their accounts on Telegram that there were no charges levied against the proprietors of 1337 Services GmbH. FlorainN told former customers they were in the process of moving to a new name and domain for StarkRDP, where existing accounts and balances would be transferred.

“StarkRDP has always been operating by the law and is not involved in any of these alleged crimes and the legal process will confirm this,” the StarkRDP Telegram account wrote on January 30. “All of your servers are safe and they have not been collected in this operation. The only things that were seized is the website server and our domain. Unfortunately, no one can tell who took it and with whom we can talk about it. Therefore, we will restart operation soon, under a different name, to close the chapter [of] ‘StarkRDP.'”

FBI, Dutch Police Disrupt ‘Manipulaters’ Phishing Gang

The FBI and authorities in The Netherlands this week seized dozens of servers and domains for a hugely popular spam and malware dissemination service operating out of Pakistan. The proprietors of the service, who use the collective nickname “The Manipulaters,” have been the subject of three stories published here since 2015. The FBI said the main clientele are organized crime groups that try to trick victim companies into making payments to a third party.

On January 29, the FBI and the Dutch national police seized the technical infrastructure for a cybercrime service marketed under the brands Heartsender, Fudpage and Fudtools (and many other “fud” variations). The “fud” bit stands for “Fully Un-Detectable,” and it refers to cybercrime resources that will evade detection by security tools like antivirus software or anti-spam appliances.

The Dutch authorities said 39 servers and domains abroad were seized, and that the servers contained millions of records from victims worldwide — including at least 100,000 records pertaining to Dutch citizens.

A statement from the U.S. Department of Justice refers to the cybercrime group as Saim Raza, after a pseudonym The Manipulaters communally used to promote their spam, malware and phishing services on social media.

“The Saim Raza-run websites operated as marketplaces that advertised and facilitated the sale of tools such as phishing kits, scam pages and email extractors often used to build and maintain fraud operations,” the DOJ explained.

The core Manipulaters product is Heartsender, a spam delivery service whose homepage openly advertised phishing kits targeting users of various Internet companies, including Microsoft 365, Yahoo, AOL, Intuit, iCloud and ID.me, to name a few.

The government says transnational organized crime groups that purchased these services primarily used them to run business email compromise (BEC) schemes, wherein the cybercrime actors tricked victim companies into making payments to a third party.

“Those payments would instead be redirected to a financial account the perpetrators controlled, resulting in significant losses to victims,” the DOJ wrote. “These tools were also used to acquire victim user credentials and utilize those credentials to further these fraudulent schemes. The seizure of these domains is intended to disrupt the ongoing activity of these groups and stop the proliferation of these tools within the cybercriminal community.”

KrebsOnSecurity first wrote about The Manipulaters in May 2015, mainly because their ads at the time were blanketing a number of popular cybercrime forums, and because they were fairly open and brazen about what they were doing — even who they were in real life.

We caught up with The Manipulaters again in 2021, with a story that found the core employees had started a web coding company in Lahore called WeCodeSolutions — presumably as a way to account for their considerable Heartsender income. That piece examined how WeCodeSolutions employees had all doxed themselves on Facebook by posting pictures from company parties each year featuring a large cake with the words FudCo written in icing.

A follow-up story last year about The Manipulaters prompted messages from various WeCodeSolutions employees who pleaded with this publication to remove stories about them. The Saim Raza identity told KrebsOnSecurity they were recently released from jail after being arrested and charged by local police, although they declined to elaborate on the charges.

The Manipulaters never seemed to care much about protecting their own identities, so it’s not surprising that they were unable or unwilling to protect their own customers. In an analysis released last year, DomainTools.com found the web-hosted version of Heartsender leaked an extraordinary amount of user information to unauthenticated users, including customer credentials and email records from Heartsender employees.

DomainTools also uncovered evidence that the computers used by The Manipulaters were all infected with the same password-stealing malware, and that vast numbers of credentials were stolen from the group and sold online.

“Ironically, the Manipulaters may create more short-term risk to their own customers than law enforcement,” DomainTools wrote. “The data table ‘User Feedbacks’ (sic) exposes what appear to be customer authentication tokens, user identifiers, and even a customer support request that exposes root-level SMTP credentials–all visible by an unauthenticated user on a Manipulaters-controlled domain.”

Police in The Netherlands said the investigation into the owners and customers of the service is ongoing.

“The Cybercrime Team is on the trail of a number of buyers of the tools,” the Dutch national police said. “Presumably, these buyers also include Dutch nationals. The investigation into the makers and buyers of this phishing software has not yet been completed with the seizure of the servers and domains.”

U.S. authorities this week also joined law enforcement in Australia, France, Greece, Italy, Romania and Spain in seizing a number of domains for several long-running cybercrime forums and services, including Cracked and Nulled. According to a statement from the European police agency Europol, the two communities attracted more than 10 million users in total.

Other domains seized as part of “Operation Talent” included Sellix, an e-commerce platform that was frequently used by cybercrime forum members to buy and sell illicit goods and services.

How Cryptocurrency Turns to Cash in Russian Banks

A financial firm registered in Canada has emerged as the payment processor for dozens of Russian cryptocurrency exchanges and websites hawking cybercrime services aimed at Russian-speaking customers, new research finds. Meanwhile, an investigation into the Vancouver street address used by this company shows it is home to dozens of foreign currency dealers, money transfer businesses, and cryptocurrency exchanges — none of which are physically located there.

Richard Sanders is a blockchain analyst and investigator who advises the law enforcement and intelligence community. Sanders spent most of 2023 in Ukraine, traveling with Ukrainian soldiers while mapping the shifting landscape of Russian crypto exchanges that are laundering money for narcotics networks operating in the region.

More recently, Sanders has focused on identifying how dozens of popular cybercrime services are getting paid by their customers, and how they are converting cryptocurrency revenues into cash. For the past several months, he’s been signing up for various cybercrime services, and then tracking where their customer funds go from there.

The 122 services targeted in Sanders’ research include some of the more prominent businesses advertising on the cybercrime forums today, such as:

-abuse-friendly or “bulletproof” hosting providers like anonvm[.]wtf, and PQHosting;
-sites selling aged email, financial, or social media accounts, such as verif[.]work and kopeechka[.]store;
-anonymity or “proxy” providers like crazyrdp[.]com and rdp[.]monster;
-anonymous SMS services, including anonsim[.]net and smsboss[.]pro.

Sanders said he first encountered some of these services while investigating Kremlin-funded disinformation efforts in Ukraine, as they are all useful in assembling large-scale, anonymous social media campaigns.

According to Sanders, all 122 of the services he tested are processing transactions through a company called Cryptomus, which says it is a cryptocurrency payments platform based in Vancouver, British Columbia. Cryptomus’ website says its parent firm — Xeltox Enterprises Ltd. (formerly certa-pay[.]com) — is registered as a money service business (MSB) with the Financial Transactions and Reports Analysis Centre of Canada (FINTRAC).

Sanders said the payment data he gathered also shows that at least 56 cryptocurrency exchanges are currently using Cryptomus to process transactions, including financial entities with names like casher[.]su, grumbot[.]com, flymoney[.]biz, obama[.]ru and swop[.]is.

These platforms are built for Russian speakers, and they each advertise the ability to anonymously swap one form of cryptocurrency for another. They also allow the exchange of cryptocurrency for cash in accounts at some of Russia’s largest banks — nearly all of which are currently sanctioned by the United States and other western nations.

An analysis of their technology infrastructure shows that all of these exchanges use Russian email providers, and most are directly hosted in Russia or by Russia-backed ISPs with infrastructure in Europe (e.g. Selectel, Netwarm UK, Beget, Timeweb and DDoS-Guard). The analysis also showed nearly all 56 exchanges used services from Cloudflare, a global content delivery network based in San Francisco.

“Purportedly, the purpose of these platforms is for companies to accept cryptocurrency payments in exchange for goods or services,” Sanders told KrebsOnSecurity. “Unfortunately, it is next to impossible to find any goods for sale with websites using Cryptomus, and the services appear to fall into one or two different categories: Facilitating transactions with sanctioned Russian banks, and platforms providing the infrastructure and means for cyber attacks.”

Cryptomus did not respond to multiple requests for comment.

PHANTOM ADDRESSES?

The Cryptomus website and its FINTRAC listing say the company’s registered address is Suite 170, 422 Richards St. in Vancouver, BC. This address was the subject of an investigation published in July by CTV National News and the Investigative Journalism Foundation (IJF), which documented dozens of cases across Canada where multiple MSBs are incorporated at the same address, often without the knowledge or consent of the location’s actual occupant.

Their inquiry found 422 Richards St. was listed as the registered address for at least 76 foreign currency dealers, eight MSBs, and six cryptocurrency exchanges. At that address is a three-story building that used to be a bank and now houses a massage therapy clinic and a co-working space. But they found none of the MSBs or currency dealers were paying for services at that co-working space.

The reporters found another collection of 97 MSBs clustered at an address for a commercial office suite in Ontario, even though there was no evidence these companies had ever arranged for any business services at that address.

Peter German, a former deputy commissioner for the Royal Canadian Mounted Police who authored two reports on money laundering in British Columbia, told the publications it goes against the spirit of Canada’s registration requirements for such businesses, which are considered high-risk for money laundering and terrorist financing.

“If you’re able to have 70 in one building, that’s just an abuse of the whole system,” German said.

Ten MSBs registered to 422 Richard St. had their registrations revoked. One company at 422 Richards St. whose registration was revoked this year had a director with a listed address in Russia, the publications reported. “Others appear to be directed by people who are also directors of companies in Cyprus and other high-risk jurisdictions for money laundering,” they wrote.

A review of FINTRAC’s registry (.CSV) shows many of the MSBs at 422 Richards St. are international money transfer or remittance services to countries like Malaysia, India and Nigeria. Some act as currency exchanges, while others appear to sell merchant accounts and online payment services. Still, KrebsOnSecurity could find no obvious connections between the 56 Russian cryptocurrency exchanges identified by Sanders and the dozens of payment companies that FINTRAC says share an address with the Cryptomus parent firm Xeltox Enterprises.

SANCTIONS EVASION

In August 2023, Binance and some of the largest cryptocurrency exchanges responded to sanctions against Russia by cutting off many Russian banks and restricting Russian customers to transactions in Rubles only. Sanders said prior to that change, most of the exchanges currently served by Cryptomus were handling customer funds with their own self-custodial cryptocurrency wallets.

By September 2023, Sanders said he found the exchanges he was tracking had all nested themselves like Matryoshka dolls at Cryptomus, which adds a layer of obfuscation to all transactions by generating a new cryptocurrency wallet for each order.

“They all simply moved to Cryptomus,” he said. “Cryptomus generates new wallets for each order, rendering ongoing attribution to require transactions with high fees each time.”

“Exchanges like Binance and OKX removing Sberbank and other sanctioned banks and offboarding Russian users did not remove the ability of Russians to transact in and out of cryptocurrency easily,” he continued. “In fact, it’s become easier, because the instant-swap exchanges do not even have Know Your Customer rules. The U.S. sanctions resulted in the majority of Russian instant exchanges switching from their self-custodial wallets to platforms, especially Cryptomus.”

Russian President Vladimir Putin in August signed a new law legalizing cryptocurrency mining and allowing the use of cryptocurrency for international payments. The Russian government’s embrace of cryptocurrency was a remarkable pivot: Bloomberg notes that as recently as January 2022, just weeks before Russia’s full-scale invasion of Ukraine, the central bank proposed a blanket ban on the use and creation of cryptocurrencies.

In a report on Russia’s cryptocurrency ambitions published in September, blockchain analysis firm Chainalysis said Russia’s move to integrate crypto into its financial system may improve its ability to bypass the U.S.-led financial system and to engage in non-dollar denominated trade.

“Although it can be hard to quantify the true impact of certain sanctions actions, the fact that Russian officials have singled out the effect of sanctions on Moscow’s ability to process cross-border trade suggests that the impact felt is great enough to incite urgency to legitimize and invest in alternative payment channels it once decried,” Chainalysis assessed.

Asked about its view of activity on Cryptomus, Chainanlysis said Cryptomus has been used by criminals of all stripes for laundering money and/or the purchase of goods and services.

“We see threat actors engaged in ransomware, narcotics, darknet markets, fraud, cybercrime, sanctioned entities and jurisdictions, and hacktivism making deposits to Cryptomus for purchases but also laundering the services using Cryptomos payment API,” the company said in a statement.

SHELL GAMES

It is unclear if Cryptomus and/or Xeltox Enterprises have any presence in Canada at all. A search in the United Kingdom’s Companies House registry for Xeltox’s former name — Certa Payments Ltd. — shows an entity by that name incorporated at a mail drop in London in December 2023.

The sole shareholder and director of that company is listed as a 25-year-old Ukrainian woman in the Czech Republic named Vira Krychka. Ms. Krychka was recently appointed the director of several other new U.K. firms, including an entity created in February 2024 called Globopay UAB Ltd, and another called WS Management and Advisory Corporation Ltd. Ms. Krychka did not respond to a request for comment.

WS Management and Advisory Corporation bills itself as the regulatory body that exclusively oversees licenses of cryptocurrencies in the jurisdiction of Western Sahara, a disputed territory in northwest Africa. Its website says the company assists applicants with bank setup and formation, online gaming licenses, and the creation and licensing of foreign exchange brokers. One of Certa Payments’ former websites — certa[.]website — also shared a server with 12 other domains, including rasd-state[.]ws, a website for the Central Reserve Authority of the Western Sahara.

This business registry from the Czech Republic indicates Ms. Krychka works as a director at an advertising and marketing firm called Icon Tech SRO, which was previously named Blaven Technologies (Blaven’s website says it is an online payment service provider).

In August 2024, Icon Tech changed its name again to Mezhundarondnaya IBU SRO, which describes itself as an “experienced company in IT consulting” that is based in Armenia. The same registry says Ms. Krychka is somehow also a director at a Turkish investment venture. So much business acumen at such a young age!

For now, Canada remains an attractive location for cryptocurrency businesses to set up shop, at least on paper. The IJF and CTV News found that as of February 2024, there were just over 3,000 actively registered MSBs in Canada, 1,247 of which were located at the same building as at least one other MSB.

“That analysis does not include the roughly 2,700 MSBs whose registrations have lapsed, been revoked or otherwise stopped,” they observed. “If they are included, then a staggering 2,061 out of 5,705 total MSBs share a building with at least one other MSB.”

Why Phishers Love New TLDs Like .shop, .top and .xyz

Phishing attacks increased nearly 40 percent in the year ending August 2024, with much of that growth concentrated at a small number of new generic top-level domains (gTLDs) — such as .shop, .top, .xyz — that attract scammers with rock-bottom prices and no meaningful registration requirements, new research finds. Meanwhile, the nonprofit entity that oversees the domain name industry is moving forward with plans to introduce a slew of new gTLDs.

A study on phishing data released by Interisle Consulting finds that new gTLDs introduced in the last few years command just 11 percent of the market for new domains, but accounted for roughly 37 percent of cybercrime domains reported between September 2023 and August 2024.

Interisle was sponsored by several anti-spam organizations, including the Anti-Phishing Working Group (APWG), the Coalition Against Unsolicited Commercial Email (CAUCE), and the Messaging, Malware, and Mobile Anti-Abuse Working Group (M3AAWG).

The study finds that while .com and .net domains made up approximately half of all domains registered in the past year (more than all of the other TLDs combined) they accounted for just over 40 percent of all cybercrime domains. Interisle says an almost equal share — 37 percent — of cybercrime domains were registered through new gTLDs.

Spammers and scammers gravitate toward domains in the new gTLDs because these registrars tend to offer cheap or free registration with little to no account or identity verification requirements. For example, among the gTLDs with the highest cybercrime domain scores in this year’s study, nine offered registration fees for less than $1, and nearly two dozen offered fees of less than $2.00. By comparison, the cheapest price identified for a .com domain was $5.91.

Currently, there are around 2,500 registrars authorized to sell domains by the Internet Corporation for Assigned Names and Numbers (ICANN), the California nonprofit that oversees the domain industry.

Incredibly, despite years of these reports showing phishers heavily abusing new gTLDs, ICANN is shuffling forward on a plan to introduce even more of them. ICANN’s proposed next round envisions accepting applications for new gTLDs in 2026.

John Levine is author of the book “The Internet for Dummies” and president of CAUCE. Levine said adding more TLDs without a much stricter registration policy will likely further expand an already plentiful greenfield for cybercriminals.

“The problem is that ICANN can’t make up their mind whether they are the neutral nonprofit regulator or just the domain speculator trade association,” Levine told KrebsOnSecurity. “But they act a lot more like the latter.”

Levine said the vast majority of new gTLDs have a few thousand domains — a far cry from the number of registrations they would need just to cover the up-front costs of operating a new gTLD (~$180,000-$300,000). New gTLD registrars can quickly attract customers by selling domains cheaply to customers who buy domains in bulk, but that tends to be a losing strategy.

“Selling to criminals and spammers turns out to be lousy business,” Levine said. “You can charge whatever you want on the first year, but you have to charge list price on domain renewals. And criminals and spammers never renew. So if it sounds like the economics makes no sense it’s because the economics makes no sense.”

In virtually all previous spam reports, Interisle found the top brands referenced in phishing attacks were the largest technology companies, including Apple, Facebook, Google and PayPal. But this past year, Interisle found the U.S. Postal Service was by far the most-phished entity, with more than four times the number of phishing domains as the second most-frequent target (Apple).

At least some of that increase is likely from a prolific cybercriminal using the nickname Chenlun, who has been selling phishing kits targeting domestic postal services in the United States and at least a dozen other countries.

Interisle says an increasing number of phishers are eschewing domain registrations altogether, and instead taking advantage of subdomain providers like blogspot.com, pages.dev, and weebly.com. The report notes that cyberattacks hosted at subdomain provider services can be tough to mitigate, because only the subdomain provider can disable malicious accounts or take down malicious web pages.

“Any action upstream, such as blocking the second-level domain, would have an impact across the provider’s whole customer base,” the report observes.

Interisle tracked more than 1.18 million instances of subdomains used for phishing in the past year (a 114 percent increase), and found more than half of those were subdomains at blogspot.com and other services operated by Google.

“Many of these services allow the creation of large numbers of accounts at one time, which is highly exploited by criminals,” the report concludes. “Subdomain providers should limit the number of subdomains (user accounts) a customer can create at one time and suspend automated, high-volume automated account sign-ups – especially using free services.”

Dec. 4, 10:21 a.m. ET: Corrected link to report.

Canadian Man Arrested in Snowflake Data Extortions

A 25-year-old man in Ontario, Canada has been arrested for allegedly stealing data from and extorting more than 160 companies that used the cloud data service Snowflake.

On October 30, Canadian authorities arrested Alexander Moucka, a.k.a. Connor Riley Moucka of Kitchener, Ontario, on a provisional arrest warrant from the United States. Bloomberg first reported Moucka’s alleged ties to the Snowflake hacks on Monday.

At the end of 2023, malicious hackers learned that many large companies had uploaded huge volumes of sensitive customer data to Snowflake accounts that were protected with little more than a username and password (no multi-factor authentication required). After scouring darknet markets for stolen Snowflake account credentials, the hackers began raiding the data storage repositories used by some of the world’s largest corporations.

Among those was AT&T, which disclosed in July that cybercriminals had stolen personal information and phone and text message records for roughly 110 million people — nearly all of its customers. Wired.com reported in July that AT&T paid a hacker $370,000 to delete stolen phone records.

A report on the extortion attacks from the incident response firm Mandiant notes that Snowflake victim companies were privately approached by the hackers, who demanded a ransom in exchange for a promise not to sell or leak the stolen data. All told, more than 160 Snowflake customers were relieved of data, including TicketMaster, Lending Tree, Advance Auto Parts and Neiman Marcus.

Moucka is alleged to have used the hacker handles Judische and Waifu, among many others. These monikers correspond to a prolific cybercriminal whose exploits were the subject of a recent story published here about the overlap between Western, English-speaking cybercriminals and extremist groups that harass and extort minors into harming themselves or others.

On May 2, 2024, Judische claimed on the fraud-focused Telegram channel Star Chat that they had hacked Santander Bank, one of the first known Snowflake victims. Judische would repeat that claim in Star Chat on May 13 — the day before Santander publicly disclosed a data breach — and would periodically blurt out the names of other Snowflake victims before their data even went up for sale on the cybercrime forums.

404 Media reports that at a court hearing in Ontario this morning, Moucka called in from a prison phone and said he was seeking legal aid to hire an attorney.

TELECOM DOMINOES

Mandiant has attributed the Snowflake compromises to a group it calls “UNC5537,” with members based in North America and Turkey. Sources close to the investigation tell KrebsOnSecurity the UNC5537 member in Turkey is John Erin Binns, an elusive American man indicted by the U.S. Department of Justice (DOJ) for a 2021 breach at T-Mobile that exposed the personal information of at least 76.6 million customers.

Update: The Justice Department has unsealed an indictment (PDF) against Moucka and Binns, charging them with one count of conspiracy; 10 counts of wire fraud; four counts of computer fraud and abuse; two counts of extortion in relation to computer fraud; and two counts aggravated identity theft.

In a statement on Moucka’s arrest, Mandiant said UNC5537 aka Alexander ‘Connor’ Moucka has proven to be one of the most consequential threat actors of 2024.

“In April 2024, UNC5537 launched a campaign, systematically compromising misconfigured SaaS instances across over a hundred organizations,” wrote Austin Larsen, Mandiant’s senior threat analyst. “The operation, which left organizations reeling from significant data loss and extortion attempts, highlighted the alarming scale of harm an individual can cause using off-the-shelf tools.”

Sources involved in the investigation said UNC5537 has focused on hacking into telecommunications companies around the world. Those sources told KrebsOnSecurity that Binns and Judische are suspected of stealing data from India’s largest state-run telecommunications firm Bharat Sanchar Nigam Ltd (BNSL), and that the duo even bragged about being able to intercept or divert phone calls and text messages for a large portion of the population of India.

Judische appears to have outsourced the sale of databases from victim companies who refuse to pay, delegating some of that work to a cybercriminal who uses the nickname Kiberphant0m on multiple forums. In late May 2024, Kiberphant0m began advertising the sale of hundreds of gigabytes of data stolen from BSNL.

“Information is worth several million dollars but I’m selling for pretty cheap,” Kiberphant0m wrote of the BSNL data in a post on the English-language cybercrime community Breach Forums. “Negotiate a deal in Telegram.”

Also in May 2024, Kiberphant0m took to the Russian-language hacking forum XSS to sell more than 250 gigabytes of data stolen from an unnamed mobile telecom provider in Asia, including a database of all active customers and software allowing the sending of text messages to all customers.

On September 3, 2024, Kiberphant0m posted a sales thread on XSS titled “Selling American Telecom Access (100B+ Revenue).” Kiberphant0m’s asking price of $200,000 was apparently too high because they reposted the sales thread on Breach Forums a month later, with a headline that more clearly explained the data was stolen from Verizon‘s “push-to-talk” (PTT) customers — primarily U.S. government agencies and first responders.

404Media reported recently that the breach does not appear to impact the main consumer Verizon network. Rather, the hackers broke into a third party provider and stole data on Verizon’s PTT systems, which are a separate product marketed towards public sector agencies, enterprises, and small businesses to communicate internally.

INTERVIEW WITH JUDISCHE

Investigators say Moucka shared a home in Kitchener with other tenants, but not his family. His mother was born in Chechnya, and he speaks Russian in addition to French and English. Moucka’s father died of a drug overdose at age 26, when the defendant was roughly five years old.

A person claiming to be Judische began communicating with this author more than three months ago on Signal after KrebsOnSecurity started asking around about hacker nicknames previously used by Judische over the years.

Judische admitted to stealing and ransoming data from Snowflake customers, but he said he’s not interested in selling the information, and that others have done this with some of the data sets he stole.

“I’m not really someone that sells data unless it’s crypto [databases] or credit cards because they’re the only thing I can find buyers for that actually have money for the data,” Judische told KrebsOnSecurity. “The rest is just ransom.”

Judische has sent this reporter dozens of unsolicited and often profane messages from several different Signal accounts, all of which claimed to be an anonymous tipster sharing different identifying details for Judische. This appears to have been an elaborate effort by Judische to “detrace” his movements online and muddy the waters about his identity.

Judische frequently claimed he had unparalleled “opsec” or operational security, a term that refers to the ability to compartmentalize and obfuscate one’s tracks online. In an effort to show he was one step ahead of investigators, Judische shared information indicating someone had given him a Mandiant researcher’s assessment of who and where they thought he was. Mandiant says those were discussion points shared with select reporters in advance of the researcher’s recent talk at the LabsCon security conference.

But in a conversation with KrebsOnSecurity on October 26, Judische acknowledged it was likely that the authorities were closing in on him, and said he would seriously answer certain questions about his personal life.

“They’re coming after me for sure,” he said.

In several previous conversations, Judische referenced suffering from an unspecified personality disorder, and when pressed said he has a condition called “schizotypal personality disorder” (STPD).

According to the Cleveland Clinic, schizotypal personality disorder is marked by a consistent pattern of intense discomfort with relationships and social interactions: “People with STPD have unusual thoughts, speech and behaviors, which usually hinder their ability to form and maintain relationships.”

Judische said he was prescribed medication for his psychological issues, but that he doesn’t take his meds. Which might explain why he never leaves his home.

“I never go outside,” Judische allowed. “I’ve never had a friend or true relationship not online nor in person. I see people as vehicles to achieve my ends no matter how friendly I may seem on the surface, which you can see by how fast I discard people who are loyal or [that] I’ve known a long time.”

Judische later admitted he doesn’t have an official STPD diagnosis from a physician, but said he knows that he exhibits all the signs of someone with this condition.

“I can’t actually get diagnosed with that either,” Judische shared. “Most countries put you on lists and restrict you from certain things if you have it.”

Asked whether he has always lived at his current residence, Judische replied that he had to leave his hometown for his own safety.

“I can’t live safely where I’m from without getting robbed or arrested,” he said, without offering more details.

A source familiar with the investigation said Moucka previously lived in Quebec, which he allegedly fled after being charged with harassing others on the social network Discord.

Judische claims to have made at least $4 million in his Snowflake extortions. Judische said he and others frequently targeted business process outsourcing (BPO) companies, staffing firms that handle customer service for a wide range of organizations. They also went after managed service providers (MSPs) that oversee IT support and security for multiple companies, he claimed.

“Snowflake isn’t even the biggest BPO/MSP multi-company dataset on our networks, but what’s been exfiltrated from them is well over 100TB,” Judische bragged. “Only ones that don’t pay get disclosed (unless they disclose it themselves). A lot of them don’t even do their SEC filing and just pay us to fuck off.”

INTEL SECRETS

The other half of UNC5537 — 24-year-old John Erin Binns — was arrested in Turkey in late May 2024, and currently resides in a Turkish prison. However, it is unclear if Binns faces any immediate threat of extradition to the United States, where he is currently wanted on criminal hacking charges tied to the 2021 breach at T-Mobile.

A person familiar with the investigation said Binns’s application for Turkish citizenship was inexplicably approved after his incarceration, leading to speculation that Binns may have bought his way out of a sticky legal situation.

Under the Turkish constitution, a Turkish citizen cannot be extradited to a foreign state. Turkey has been criticized for its “golden passport” program, which provides citizenship and sanctuary for anyone willing to pay several hundred thousand dollars.

Binns’s alleged hacker alter egos — “IRDev” and “IntelSecrets” — were at once feared and revered on several cybercrime-focused Telegram communities, because he was known to possess a powerful weapon: A massive botnet. From reviewing the Telegram channels Binns frequented, we can see that others in those communities — including Judische — heavily relied on Binns and his botnet for a variety of cybercriminal purposes.

The IntelSecrets nickname corresponds to an individual who has claimed responsibility for modifying the source code for the Mirai “Internet of Things” botnet to create a variant known as “Satori,” and supplying it to others who used it for criminal gain and were later caught and prosecuted.

Since 2020, Binns has filed a flood of lawsuits naming various federal law enforcement officers and agencies — including the FBI, the CIA, and the U.S. Special Operations Command (PDF), demanding that the government turn over information collected about him and seeking restitution for his alleged kidnapping at the hands of the CIA.

Binns claims he was kidnapped in Turkey and subjected to various forms of psychological and physical torture. According to Binns, the U.S. Central Intelligence Agency (CIA) falsely told their counterparts in Turkey that he was a supporter or member of the Islamic State (ISIS), a claim he says led to his detention and torture by the Turkish authorities.

However, in a 2020 lawsuit he filed against the CIA, Binns himself acknowledged having visited a previously ISIS-controlled area of Syria prior to moving to Turkey in 2017.

Sources familiar with the investigation told KrebsOnSecurity that Binns was so paranoid about possible surveillance on him by American and Turkish intelligence agencies that his erratic behavior and online communications actually brought about the very government snooping that he feared.

In several online chats in late 2023 on Discord, IRDev lamented being lured into a law enforcement sting operation after trying to buy a rocket launcher online. A person close to the investigation confirmed that at the beginning of 2023, IRDev began making earnest inquiries about how to purchase a Stinger, an American-made portable weapon that operates as an infrared surface-to-air missile.

Sources told KrebsOnSecurity Binns’ repeated efforts to purchase the projectile earned him multiple visits from the Turkish authorities, who were justifiably curious why he kept seeking to acquire such a powerful weapon.

WAIFU

A careful study of Judische’s postings on Telegram and Discord since 2019 shows this user is more widely known under the nickname “Waifu,” a moniker that corresponds to one of the more accomplished “SIM swappers” in the English-language cybercrime community over the years.

SIM swapping involves phishing, tricking or bribing mobile phone company employees for credentials needed to redirect a target’s mobile phone number to a device the attackers control — allowing thieves to intercept incoming text messages and phone calls.

Several SIM-swapping channels on Telegram maintain a frequently updated leaderboard of the 100 richest SIM-swappers, as well as the hacker handles associated with specific cybercrime groups (Waifu is ranked #24). That list has long included Waifu on a roster of hackers for a group that called itself “Beige.”

The term “Beige Group” came up in reporting on two stories published here in 2020. The first was in an August 2020 piece called Voice Phishers Targeting Corporate VPNs, which warned that the COVID-19 epidemic had brought a wave of targeted voice phishing attacks that tried to trick work-at-home employees into providing access to their employers’ networks. Frequent targets of the Beige group included employees at numerous top U.S. banks, ISPs, and mobile phone providers.

The second time Beige Group was mentioned by sources was in reporting on a breach at the domain registrar GoDaddy. In November 2020, intruders thought to be associated with the Beige Group tricked a GoDaddy employee into installing malicious software, and with that access they were able to redirect the web and email traffic for multiple cryptocurrency trading platforms. Other frequent targets of the Beige group included employees at numerous top U.S. banks, ISPs, and mobile phone providers.

Judische’s various Telegram identities have long claimed involvement in the 2020 GoDaddy breach, and he didn’t deny his alleged role when asked directly. Judische said he prefers voice phishing or “vishing” attacks that result in the target installing data-stealing malware, as opposed to tricking the user into entering their username, password and one-time code.

“Most of my ops involve malware [because] credential access burns too fast,” Judische explained.

CRACKDOWN ON HARM GROUPS?

The Telegram channels that the Judische/Waifu accounts frequented over the years show this user divided their time between posting in channels dedicated to financial cybercrime, and harassing and stalking others in harm communities like Leak Society and Court.

Both of these Telegram communities are known for victimizing children through coordinated online campaigns of extortion, doxing, swatting and harassment. People affiliated with harm groups like Court and Leak Society will often recruit new members by lurking on gaming platforms, social media sites and mobile applications that are popular with young people, including Discord, Minecraft, Roblox, Steam, Telegram, and Twitch.

“This type of offence usually starts with a direct message through gaming platforms and can move to more private chatrooms on other virtual platforms, typically one with video enabled features, where the conversation quickly becomes sexualized or violent,” warns a recent alert from the Royal Canadian Mounted Police (RCMP) about the rise of sextortion groups on social media channels.

“One of the tactics being used by these actors is sextortion, however, they are not using it to extract money or for sexual gratification,” the RCMP continued. “Instead they use it to further manipulate and control victims to produce more harmful and violent content as part of their ideological objectives and radicalization pathway.”

Some of the largest such known groups include those that go by the names 764, CVLT, Kaskar, 7997, 8884, 2992, 6996, 555, Slit Town, 545, 404, NMK, 303, and H3ll.

On the various cybercrime-oriented channels Judische frequented, he often lied about his or others’ involvement in various breaches. But Judische also at times shared nuggets of truth about his past, particularly when discussing the early history and membership of specific Telegram- and Discord-based cybercrime and harm groups.

Judische claimed in multiple chats, including on Leak Society and Court, that they were an early member of the Atomwaffen Division (AWD), a white supremacy group whose members are suspected of having committed multiple murders in the U.S. since 2017.

In 2019, KrebsOnSecurity exposed how a loose-knit group of neo-Nazis, some of whom were affiliated with AWD, had doxed and/or swatted nearly three dozen journalists at a range of media publications. Swatting involves communicating a false police report of a bomb threat or hostage situation and tricking authorities into sending a heavily armed police response to a targeted address.

Judsiche also told a fellow denizen of Court that years ago he was active in an older harm community called “RapeLash,” a truly vile Discord server known for attracting Atomwaffen members. A 2018 retrospective on RapeLash posted to the now defunct neo-Nazi forum Fascist Forge explains that RapeLash was awash in gory, violent images and child pornography.

A Fascist Forge member named “Huddy” recalled that RapeLash was the third incarnation of an extremist community also known as “FashWave,” short for Fascist Wave.

“I have no real knowledge of what happened with the intermediary phase known as ‘FashWave 2.0,’ but FashWave 3.0 houses multiple known Satanists and other degenerates connected with AWD, one of which got arrested on possession of child pornography charges, last I heard,” Huddy shared.

In June 2024, a Mandiant employee told Bloomberg that UNC5537 members have made death threats against cybersecurity experts investigating the hackers, and that in one case the group used artificial intelligence to create fake nude photos of a researcher to harass them.

Allison Nixon is chief research officer with the New York-based cybersecurity firm Unit 221B. Nixon is among several researchers who have faced harassment and specific threats of physical violence from Judische.

Nixon said Judische is likely to argue in court that his self-described psychological disorder(s) should somehow excuse his long career in cybercrime and in harming others.

“They ran a misinformation campaign in a sloppy attempt to cover up the hacking campaign,” Nixon said of Judische. “Coverups are an acknowledgment of guilt, which will undermine a mental illness defense in court. We expect that violent hackers from the [cybercrime community] will experience increasingly harsh sentences as the crackdown continues.”

5:34 p.m. ET: Updated story to include a clarification from Mandiant. Corrected Moucka’s age.

Nov. 21, 2024: Included link to a criminal indictment against Moucka and Binns.

Low-Drama ‘Dark Angels’ Reap Record Ransoms

A ransomware group called Dark Angels made headlines this past week when it was revealed the crime group recently received a record $75 million data ransom payment from a Fortune 50 company. Security experts say the Dark Angels have been around since 2021, but the group doesn’t get much press because they work alone and maintain a low profile, picking one target at a time and favoring mass data theft over disrupting the victim’s operations.

Security firm Zscaler ThreatLabz this month ranked Dark Angels as the top ransomware threat for 2024, noting that in early 2024 a victim paid the ransomware group $75 million — higher than any previously recorded ransom payment. ThreatLabz found Dark Angels has conducted some of the largest ransomware attacks to date, and yet little is known about the group.

Brett Stone-Gross, senior director of threat intelligence at ThreatLabz, said Dark Angels operate using an entirely different playbook than most other ransomware groups. For starters, he said, Dark Angels does not employ the typical ransomware affiliate model, which relies on hackers-for-hire to install malicious software that locks up infected systems.

“They really don’t want to be in the headlines or cause business disruptions,” Stone-Gross said. “They’re about making money and attracting as little attention as possible.”

Most ransomware groups maintain flashy victim leak sites which threaten to publish the target’s stolen data unless a ransom demand is paid. But the Dark Angels didn’t even have a victim shaming site until April 2023. And the leak site isn’t particularly well branded; it’s called Dunghill Leak.

“Nothing about them is flashy,” Stone-Gross said. “For the longest time, they didn’t even want to cause a big headline, but they probably felt compelled to create that leaks site because they wanted to show they were serious and that they were going to post victim data and make it accessible.”

Dark Angels is thought to be a Russia-based cybercrime syndicate whose distinguishing characteristic is stealing truly staggering amounts of data from major companies across multiple sectors, including healthcare, finance, government and education. For large businesses, the group has exfiltrated between 10-100 terabytes of data, which can take days or weeks to transfer, ThreatLabz found.

Like most ransom gangs, Dark Angels will publish data stolen from victims who do not pay. Some of the more notable victims listed on Dunghill Leak include the global food distribution firm Sysco, which disclosed a ransomware attack in May 2023; and the travel booking giant Sabre, which was hit by the Dark Angels in September 2023.

Stone-Gross said Dark Angels is often reluctant to deploy ransomware malware because such attacks work by locking up the target’s IT infrastructure, which typically causes the victim’s business to grind to a halt for days, weeks or even months on end. And those types of breaches tend to make headlines quickly.

“They selectively choose whether they want to deploy ransomware or not,” he said. “If they deem they can encrypt some files that won’t cause major disruptions — but will give them a ton of data — that’s what they’ll do. But really, what separates them from the rest is the volume of data they’re stealing. It’s a whole order of magnitude greater with Dark Angels. Companies losing vast amounts of data will pay these high ransoms.”

So who paid the record $75 million ransom? Bleeping Computer posited on July 30 that the victim was the pharmaceutical giant Cencora (formerly AmeriSourceBergen Corporation), which reported a data security incident to the U.S. Securities and Exchange Commission (SEC) on February 21, 2024.

The SEC requires publicly-traded companies to disclose a potentially material cybersecurity event within four days of the incident. Cencora is currently #10 on the Fortune 500 list, generating more than $262 billion in revenue last year.

Cencora did not respond to questions about whether it had made a ransom payment in connection with the February cybersecurity incident, and referred KrebsOnSecurity to expenses listed under “Other” in the restructuring section of their latest quarterly financial report (PDF). That report states that the majority of the $30 million cost in “Other” was associated with the breach.

Cencora’s quarterly statement said the incident affected a standalone legacy information technology platform in one country and the foreign business unit’s ability to operate in that country for approximately two weeks.

In its most recent State of Ransomware report (PDF), security firm Sophos found the average ransomware payment had increased fivefold in the past year, from $400,000 in 2023 to $2 million. Sophos says that in more than four-fifths (82%) of cases funding for the ransom came from multiple sources. Overall, 40% of total ransom funding came from the organizations themselves and 23% from insurance providers.

Further reading: ThreatLabz ransomware report (PDF).

Phish-Friendly Domain Registry “.top” Put on Notice

The Chinese company in charge of handing out domain names ending in “.top” has been given until mid-August 2024 to show that it has put in place systems for managing phishing reports and suspending abusive domains, or else forfeit its license to sell domains. The warning comes amid the release of new findings that .top was the most common suffix in phishing websites over the past year, second only to domains ending in “.com.”

On July 16, the Internet Corporation for Assigned Names and Numbers (ICANN) sent a letter to the owners of the .top domain registry. ICANN has filed hundreds of enforcement actions against domain registrars over the years, but in this case ICANN singled out a domain registry responsible for maintaining an entire top-level domain (TLD).

Among other reasons, the missive chided the registry for failing to respond to reports about phishing attacks involving .top domains.

“Based on the information and records gathered through several weeks, it was determined that .TOP Registry does not have a process in place to promptly, comprehensively, and reasonably investigate and act on reports of DNS Abuse,” the ICANN letter reads (PDF).

ICANN’s warning redacted the name of the recipient, but records show the .top registry is operated by a Chinese entity called Jiangsu Bangning Science & Technology Co. Ltd. Representatives for the company have not responded to requests for comment.

Domains ending in .top were represented prominently in a new phishing report released today by the Interisle Consulting Group, which sources phishing data from several places, including the Anti-Phishing Working Group (APWG), OpenPhish, PhishTank, and Spamhaus.

Interisle’s newest study examined nearly two million phishing attacks in the last year, and found that phishing sites accounted for more than four percent of all new .top domains between May 2023 and April 2024. Interisle said .top has roughly 2.76 million domains in its stable, and that more than 117,000 of those were phishing sites in the past year.

ICANN said its review was based on information collected and studied about .top domains over the past few weeks. But the fact that high volumes of phishing sites are being registered through Jiangsu Bangning Science & Technology Co Ltd. is hardly a new trend.

For example, more than 10 years ago the same Chinese registrar was the fourth most common source of phishing websites, as tracked by the APWG. Bear in mind that the APWG report excerpted below was published more than a year before Jiangsu Bangning received ICANN approval to introduce and administer the new .top registry.

A fascinating new wrinkle in the phishing landscape is the growth in scam pages hosted via the InterPlanetary File System (IPFS), a decentralized data storage and delivery network that is based on peer-to-peer networking. According to Interisle, the use of IPFS to host and launch phishing attacks — which can make phishing sites more difficult to take down — increased a staggering 1,300 percent, to roughly 19,000 phishing sites reported in the last year.

Last year’s report from Interisle found that domain names ending in “.us” — the top-level domain for the United States — were among the most prevalent in phishing scams. While .us domains are not even on the Top 20 list of this year’s study, “.com” maintained its perennial #1 spot as the largest source of phishing domains overall.

A year ago, the phishiest domain registrar by far was Freenom, a now-defunct registrar that handed out free domains in several country-code TLDs, including .tk, .ml, .ga and .cf. Freenom went out of business after being sued by Meta, which alleged Freenom ignored abuse complaints while monetizing traffic to abusive domains.

Following Freenom’s demise, phishers quickly migrated to other new low-cost TLDs and to services that allow anonymous, free domain registrations — particularly subdomain services. For example, Interisle found phishing attacks involving websites created on Google’s blogspot.com skyrocketed last year more than 230 percent. Other subdomain services that saw a substantial growth in domains registered by phishers include weebly.com, github.io, wix.com, and ChangeIP, the report notes.

Interisle Consulting partner Dave Piscitello said ICANN could easily send similar warning letters to at least a half-dozen other top-level domain registries, noting that spammers and phishers tend to cycle through the same TLDs periodically — including .xyz, .info, .support and .lol, all of which saw considerably more business from phishers after Freenom’s implosion.

Piscitello said domain registrars and registries could significantly reduce the number of phishing sites registered through their services just by flagging customers who try to register huge volumes of domains at once. Their study found that at least 27% of the domains used for phishing were registered in bulk — i.e. the same registrant paid for hundreds or thousands of domains in quick succession.

The report includes a case study in which a phisher this year registered 17,562 domains over the course of an eight-hour period — roughly 38 domains per minute — using .lol domains that were all composed of random letters.

ICANN tries to resolve contract disputes privately with the registry and registrar community, and experts say the nonprofit organization usually only publishes enforcement letters when the recipient is ignoring its private notices. Indeed, ICANN’s letter notes Jiangsu Bangning didn’t even open its emailed notifications. It also cited the registry for falling behind in its ICANN membership fees.

With that in mind, a review of ICANN’s public enforcement activity suggests two trends: One is that there have been far fewer public compliance and enforcement actions in recent years — even as the number of new TLDs has expanded dramatically.

The second is that in a majority of cases, the failure of a registry or registrar to pay its annual ICANN membership fees was cited as a reason for a warning letter. A review of nearly two dozen enforcement letters ICANN has sent to domain registrars since 2022 shows that failure to pay dues was cited as a reason (or the reason) for the violation at least 75 percent of the time.

Piscitello, a former vice president of security at ICANN, said nearly all breach notices sent out while he was at ICANN were because the registrar owed money.

“I think the rest is just lipstick to suggest that ICANN’s on top of DNS Abuse,” Piscitello said.

KrebsOnSecurity has sought comment from ICANN and will update this story if they respond.

ICANN said most of its investigations are resolved and closed through the initial informal resolution stage, and that hundreds of enforcement cases are initiated during this stage with the contracted parties who are required to demonstrate compliance, become compliant, and/or present and implement remediation plans to prevent the recurrence of those enforcement issues.

“It is important to take into account that, prior to issuing any notice of breach to a registrar or registry operator, ICANN Compliance conducts an overall contractual compliance ‘health check’ of the relevant contracted party,” ICANN said in a written response to questions. “During this check, ICANN Compliance proactively reviews the contracted party’s compliance with obligations across the agreements and policies. Any additional contractual violation found during these checks is added to the Notice of Breach. It is not uncommon for parties who failed to comply with contractual obligations (whether they are related to DNS Abuse, RDDS, or others) to also be in arrears with ICANN fees.”

Update, 11:49 p.m. ET: Added statement from ICANN. Clarified Piscitello’s former role at ICANN.

The Not-So-Secret Network Access Broker x999xx

Most accomplished cybercriminals go out of their way to separate their real names from their hacker handles. But among certain old-school Russian hackers it is not uncommon to find major players who have done little to prevent people from figuring out who they are in real life. A case study in this phenomenon is “x999xx,” the nickname chosen by a venerated Russian hacker who specializes in providing the initial network access to various ransomware groups.

x999xx is a well-known “access broker” who frequently sells access to hacked corporate networks — usually in the form of remote access credentials — as well as compromised databases containing large amounts of personal and financial data.

In an analysis published in February 2019, cyber intelligence firm Flashpoint called x999xx one of the most senior and prolific members of the top-tier Russian-language cybercrime forum Exploit, where x999xx could be seen frequently advertising the sale of stolen databases and network credentials.

In August 2023, x999xx sold access to a company that develops software for the real estate industry. In July 2023, x999xx advertised the sale of Social Security numbers, names, and birthdays for the citizenry of an entire U.S. state (unnamed in the auction).

A month earlier, x999xx posted a sales thread for 80 databases taken from Australia’s largest retail company. “You may use this data to demand a ransom or do something different with it,” x999xx wrote on Exploit. “Unfortunately, the flaw was patched fast. [+] no one has used the data yet [+] the data hasn’t been used to send spam [+] the data is waiting for its time.”

In October 2022, x999xx sold administrative access to a U.S. healthcare provider.

ALIAS: MAXNM

The oldest account by the name x999xx appeared in 2009 on the Russian language cybercrime forum Verified, under the email address maxnm@ozersk.com. Ozersk is a city in the Chelyabinsk region of west-central Russia.

According to the breach tracking service Constella Intelligence, the address maxnm@ozersk.com was used more than a decade ago to create an account at Vktontakte (the Russian answer to Facebook) under the name Maxim Kirtsov from Ozersk. Mr. Kirtsov’s profile — “maxnm” — says his birthday is September 5, 1991.

The user x999xx registered on the Russian language cybercrime community Zloy in 2014 using the email address maxnmalias-1@yahoo.com. Constella says this email address was used in 2022 at the Russian shipping service cdek.ru by a Maksim Georgievich Kirtsov from Ozersk.

Additional searches on these contact details reveal that prior to 2009, x999xx favored the handle Maxnm on Russian cybercrime forums. Cyber intelligence company Intel 471 finds the user Maxnm registered on Zloy in 2006 from an Internet address in Chelyabinsk, using the email address kirtsov@telecom.ozersk.ru.

That same email address was used to create Maxnm accounts on several other crime forums, including Spamdot and Exploit in 2005 (also from Chelyabinsk), and Damagelab in 2006.

A search in Constella for the Russian version of Kirtsov’s full name — Кирцов Максим Георгиевич — brings up multiple accounts registered to maksya@icloud.com.

A review of the digital footprint for maksya@icloud.com at osint.industries reveals this address was used a decade ago to register a still-active account at imageshack.com under the name x999xx. That account features numerous screenshots of financial statements from various banks, chat logs with other hackers, and even hacked websites.

Some of the photos in that Imageshack account also appear on Kirtsov’s Vkontakte page, including images of vehicles he owns, as well as pictures of potted marijuana plants. Kirtsov’s Vkontakte profile says that in 2012 he was a faculty member of the Ozersk Technological Institute National Research Nuclear University.

The Vkontakte page lists Kirtsov’s occupation as a website called ozersk[.]today, which on the surface appears to be a blog about life in Ozersk. However, in 2019 the security firm Recorded Future published a blog post which found this domain was being used to host a malicious Cobalt Strike server.

Cobalt Strike is a commercial network penetration testing and reconnaissance tool that is sold only to vetted partners. But stolen or ill-gotten Cobalt Strike licenses are frequently abused by cybercriminal gangs to help lay the groundwork for the installation of ransomware on a victim network.

In August 2023, x999xx posted a message on Exploit saying he was interested in buying a licensed version of Cobalt Strike. A month earlier, x999xx filed a complaint on Exploit against another forum member named Cobaltforce, an apparent onetime partner whose sudden and prolonged disappearance from the community left x999xx and others in the lurch. Cobaltforce recruited people experienced in using Cobalt Strike for ransomware operations, and offered to monetize access to hacked networks for a share of the profits.

DomainTools.com finds ozersk[.]today was registered to the email address dashin2008@yahoo.com, which also was used to register roughly two dozen other domains, including x999xx[.]biz. Virtually all of those domains were registered to Maxim Kirtsov from Ozersk. Below is a mind map used to track the identities mentioned in this story.

x999xx is a prolific member of the Russian webmaster forum “Gofuckbiz,” with more than 2,000 posts over nearly a decade, according to Intel 471. In one post from 2016, x999xx asked whether anyone knew where he could buy a heat lamp that simulates sunlight, explaining that one his pet rabbits had recently perished for lack of adequate light and heat. Mr. Kirtsov’s Vkontakte page includes several pictures of caged rabbits from 2015 and earlier.

CONFIRMATION

Reached via email, Mr. Kirtsov acknowledged that he is x999xx. Kirtsov said he and his team are also regular readers of KrebsOnSecurity.

“We’re glad to hear and read you,” Kirtsov replied.

Asked whether he was concerned about the legal and moral implications of his work, Kirtsov downplayed his role in ransomware intrusions, saying he was more focused on harvesting data.

“I consider myself as committed to ethical practices as you are,” Kirtsov wrote. “I have also embarked on research and am currently mentoring students. You may have noticed my activities on a forum, which I assume you know of through information gathered from public sources, possibly using the new tool you reviewed.”

“Regarding my posts about selling access, I must honestly admit, upon reviewing my own actions, I recall such mentions but believe they were never actualized,” he continued. “Many use the forum for self-serving purposes, which explains why listings of targets for sale have dwindled — they simply ceased being viable.”

Kirtsov asserted that he is not interested in harming healthcare institutions, just in stealing their data.

“As for health-related matters, I was once acquainted with affluent webmasters who would pay up to $50 for every 1000 health-themed emails,” Kirtsov said. “Therefore, I had no interest in the more sensitive data from medical institutions like X-rays, insurance numbers, or even names; I focused solely on emails. I am proficient in SQL, hence my ease with handling data like IDs and emails. And i never doing spam or something like this.”

On the Russian crime forums, x999xx said he never targets anything or anyone in Russia, and that he has little to fear from domestic law enforcement agencies provided he remains focused on foreign adversaries.

x999xx’s lackadaisical approach to personal security mirrors that of Wazawaka, another top Russian access broker who sold access to countless organizations and even operated his own ransomware affiliate programs.

“Don’t shit where you live, travel local, and don’t go abroad,” Wazawaka said of his own personal mantra. “Mother Russia will help you. Love your country, and you will always get away with everything.”

In January 2022, KrebsOnSecurity followed clues left behind by Wazawaka to identify him as 32-year-old Mikhail Matveev from Khakassia, Russia. In May 2023, the U.S. Department of Justice indicted Matveev as a key figure in several ransomware groups that collectively extorted hundreds of millions of dollars from victim organizations. The U.S. State Department is offering a $10 million reward for information leading to the capture and/or prosecution of Matveev.

Perhaps in recognition that many top ransomware criminals are largely untouchable so long as they remain in Russia, western law enforcement agencies have begun focusing more on getting inside the heads of those individuals. These so-called “psyops” are aimed at infiltrating ransomware-as-a-service operations, disrupting major cybercrime services, and decreasing trust within cybercriminal communities.

When authorities in the U.S. and U.K. announced in February 2024 that they’d infiltrated and seized the infrastructure used by the infamous LockBit ransomware gang, they borrowed the existing design of LockBit’s victim shaming website to link instead to press releases about the takedown, and included a countdown timer that was eventually replaced with the personal details of LockBit’s alleged leader.

In May 2024, law enforcement agencies in the United States and Europe announced Operation Endgame, a coordinated action against some of the most popular cybercrime platforms for delivering ransomware and data-stealing malware. The Operation Endgame website also included a countdown timer, which served to tease the release of several animated videos that mimic the same sort of flashy, short advertisements that established cybercriminals often produce to promote their services online.

‘Operation Endgame’ Hits Malware Delivery Platforms

Law enforcement agencies in the United States and Europe today announced Operation Endgame, a coordinated action against some of the most popular cybercrime platforms for delivering ransomware and data-stealing malware. Dubbed “the largest ever operation against botnets,” the international effort is being billed as the opening salvo in an ongoing campaign targeting advanced malware “droppers” or “loaders” like IcedID, Smokeloader and Trickbot.

Operation Endgame targets the cybercrime ecosystem supporting droppers/loaders, slang terms used to describe tiny, custom-made programs designed to surreptitiously install malware onto a target system. Droppers are typically used in the initial stages of a breach, and they allow cybercriminals to bypass security measures and deploy additional harmful programs, including viruses, ransomware, or spyware.

Droppers like IcedID are most often deployed through email attachments, hacked websites, or bundled with legitimate software. For example, cybercriminals have long used paid ads on Google to trick people into installing malware disguised as popular free software, such as Microsoft Teams, Adobe Reader and Discord. In those cases, the dropper is the hidden component bundled with the legitimate software that quietly loads malware onto the user’s system.

Droppers remain such a critical, human-intensive component of nearly all major cybercrime enterprises that the most popular have turned into full-fledged cybercrime services of their own. By targeting the individuals who develop and maintain dropper services and their supporting infrastructure, authorities are hoping to disrupt multiple cybercriminal operations simultaneously.

According to a statement from the European police agency Europol, between May 27 and May 29, 2024 authorities arrested four suspects (one in Armenia and three in Ukraine), and disrupted or took down more than 100 Internet servers in Bulgaria, Canada, Germany, Lithuania, the Netherlands, Romania, Switzerland, the United Kingdom, United States and Ukraine. Authorities say they also seized more than 2,000 domain names that supported dropper infrastructure online.

In addition, Europol released information on eight fugitives suspected of involvement in dropper services and who are wanted by Germany; their names and photos were added to Europol’s “Most Wanted” list on 30 May 2024.

“It has been discovered through the investigations so far that one of the main suspects has earned at least EUR 69 million in cryptocurrency by renting out criminal infrastructure sites to deploy ransomware,” Europol wrote. “The suspect’s transactions are constantly being monitored and legal permission to seize these assets upon future actions has already been obtained.”

There have been numerous such coordinated malware takedown efforts in the past, and yet often the substantial amount of coordination required between law enforcement agencies and cybersecurity firms involved is not sustained after the initial disruption and/or arrests.

But a new website erected to detail today’s action — operation-endgame.com — makes the case that this time is different, and that more takedowns and arrests are coming. “Operation Endgame does not end today,” the site promises. “New actions will be announced on this website.”

Perhaps in recognition that many of today’s top cybercriminals reside in countries that are effectively beyond the reach of international law enforcement, actions like Operation Endgame seem increasingly focused on mind games — i.e., trolling the hackers.

Writing in this month’s issue of Wired, Matt Burgess makes the case that Western law enforcement officials have turned to psychological measures as an added way to slow down Russian hackers and cut to the heart of the sweeping cybercrime ecosystem.

“These nascent psyops include efforts to erode the limited trust the criminals have in each other, driving subtle wedges between fragile hacker egos, and sending offenders personalized messages showing they’re being watched,” Burgess wrote.

When authorities in the U.S. and U.K. announced in February 2024 that they’d infiltrated and seized the infrastructure used by the infamous LockBit ransomware gang, they borrowed the existing design of LockBit’s victim shaming website to link instead to press releases about the takedown, and included a countdown timer that was eventually replaced with the personal details of LockBit’s alleged leader.

The Operation Endgame website also includes a countdown timer, which serves to tease the release of several animated videos that mimic the same sort of flashy, short advertisements that established cybercriminals often produce to promote their services online. At least two of the videos include a substantial amount of text written in Russian.

The coordinated takedown comes on the heels of another law enforcement action this week against what the director of the FBI called “likely the world’s largest botnet ever.” On Wednesday U.S. Department of Justice (DOJ) announced the arrest of YunHe Wang, the alleged operator of the ten-year-old online anonymity service 911 S5. The government also seized 911 S5’s domains and online infrastructure, which allegedly turned computers running various “free VPN” products into Internet traffic relays that facilitated billions of dollars in online fraud and cybercrime.

Stark Industries Solutions: An Iron Hammer in the Cloud

Two weeks before Russia invaded Ukraine in February 2022, a large, mysterious new Internet hosting firm called Stark Industries Solutions materialized and quickly became the epicenter of massive distributed denial-of-service (DDoS) attacks on government and commercial targets in Ukraine and Europe. An investigation into Stark Industries reveals it is being used as a global proxy network that conceals the true source of cyberattacks and disinformation campaigns against enemies of Russia.

At least a dozen patriotic Russian hacking groups have been launching DDoS attacks since the start of the war at a variety of targets seen as opposed to Moscow. But by all accounts, few attacks from those gangs have come close to the amount of firepower wielded by a pro-Russia group calling itself “NoName057(16).”

As detailed by researchers at Radware, NoName has effectively gamified DDoS attacks, recruiting hacktivists via its Telegram channel and offering to pay people who agree to install a piece of software called DDoSia. That program allows NoName to commandeer the host computers and their Internet connections in coordinated DDoS campaigns, and DDoSia users with the most attacks can win cash prizes.

A report from the security firm Team Cymru found the DDoS attack infrastructure used in NoName campaigns is assigned to two interlinked hosting providers: MIRhosting and Stark Industries. MIRhosting is a hosting provider founded in The Netherlands in 2004. But Stark Industries Solutions Ltd was incorporated on February 10, 2022, just two weeks before the Russian invasion of Ukraine.

PROXY WARS

Security experts say that not long after the war started, Stark began hosting dozens of proxy services and free virtual private networking (VPN) services, which are designed to help users shield their Internet usage and location from prying eyes.

Proxy providers allow users to route their Internet and Web browsing traffic through someone else’s computer. From a website’s perspective, the traffic from a proxy network user appears to originate from the rented IP address, not from the proxy service customer.

These services can be used in a legitimate manner for several business purposes — such as price comparisons or sales intelligence — but they are also massively abused for hiding cybercrime activity because they can make it difficult to trace malicious traffic to its original source.

What’s more, many proxy services do not disclose how they obtain access to the proxies they are renting out, and in many cases the access is obtained through the dissemination of malicious software that turns the infected system into a traffic relay — usually unbeknownst to the legitimate owner of the Internet connection. Other proxy services will allow users to make money by renting out their Internet connection to anyone.

Spur.us is a company that tracks VPNs and proxy services worldwide. Spur finds that Stark Industries (AS44477) currently is home to at least 74 VPN services, and 40 different proxy services. As we’ll see in the final section of this story, just one of those proxy networks has over a million Internet addresses available for rent across the globe.

Raymond Dijkxhoorn operates a hosting firm in The Netherlands called Prolocation. He also co-runs SURBL, an anti-abuse service that flags domains and Internet address ranges that are strongly associated with spam and cybercrime activity, including DDoS.

Dijkxhoorn said last year SURBL heard from multiple people who said they operated VPN services whose web resources were included in SURBL’s block lists.

“We had people doing delistings at SURBL for domain names that were suspended by the registrars,” Dijkhoorn told KrebsOnSecurity. “And at least two of them explained that Stark offered them free VPN services that they were reselling.”

Dijkxhoorn added that Stark Industries also sponsored activist groups from Ukraine.

“How valuable would it be for Russia to know the real IPs from Ukraine’s tech warriors?” he observed.

CLOUDY WITH A CHANCE OF BULLETS

Richard Hummel is threat intelligence lead at NETSCOUT. Hummel said when he considers the worst of all the hosting providers out there today, Stark Industries is consistently near or at the top of that list.

“The reason is we’ve had at least a dozen service providers come to us saying, ‘There’s this network out there inundating us with traffic,'” Hummel said. “And it wasn’t even DDoS attacks. [The systems] on Stark were just scanning these providers so fast it was crashing some of their services.”

Hummel said NoName will typically launch their attacks using a mix of resources rented from major, legitimate cloud services, and those from so-called “bulletproof” hosting providers like Stark. Bulletproof providers are so named when they earn or cultivate a reputation for ignoring any abuse complaints or police reports about activity on their networks.

Combining bulletproof providers with legitimate cloud hosting, Hummel said, likely makes NoName’s DDoS campaigns more resilient because many network operators will hesitate to be too aggressive in blocking Internet addresses associated with the major cloud services.

“What we typically see here is a distribution of cloud hosting providers and bulletproof hosting providers in DDoS attacks,” he said. “They’re using public cloud hosting providers because a lot of times that’s your first layer of network defense, and because [many companies are wary of] over-blocking access to legitimate cloud resources.”

But even if the cloud provider detects abuse coming from the customer, the provider is probably not going to shut the customer down immediately, Hummel said.

“There is usually a grace period, and even if that’s only an hour or two, you can still launch a large number of attacks in that time,” he said. “And then they just keep coming back and opening new cloud accounts.”

MERCENARIES TEAM

Stark Industries is incorporated at a mail drop address in the United Kingdom. UK business records list an Ivan Vladimirovich Neculiti as the company’s secretary. Mr. Neculiti also is named as the CEO and founder of PQ Hosting Plus S.R.L. (aka Perfect Quality Hosting), a Moldovan company formed in 2019 that lists the same UK mail drop address as Stark Industries.

Reached via LinkedIn, Mr. Neculiti said PQ Hosting established Stark Industries as a “white label” of its brand so that “resellers could distribute our services using our IP addresses and their clients would not have any affairs with PQ Hosting.”

“PQ Hosting is a company with over 1,000+ of [our] own physical servers in 38 countries and we have over 100,000 clients,” he said. “Though we are not as large as Hetzner, Amazon and OVH, nevertheless we are a fast growing company that provides services to tens of thousands of private customers and legal entities.”

Asked about the constant stream of DDoS attacks whose origins have traced back to Stark Industries over the past two years, Neculiti maintained Stark hasn’t received any official abuse reports about attacks coming from its networks.

“It was probably some kind of clever attack that we did not see, I do not rule out this fact, because we have a very large number of clients and our Internet channels are quite large,” he said. “But, in this situation, unfortunately, no one contacted us to report that there was an attack from our addresses; if someone had contacted us, we would have definitely blocked the network data.”

DomainTools.com finds Ivan V. Neculiti was the owner of war[.]md, a website launched in 2008 that chronicled the history of a 1990 armed conflict in Moldova known as the Transnistria War and the Moldo-Russian war.

Transnistria is a breakaway pro-Russian region that declared itself a state in 1990, although it is not internationally recognized. The copyright on that website credits the “MercenarieS TeaM,” which was at one time a Moldovan IT firm. Mr. Neculiti confirmed personally registering this domain.

DON CHICHO & DFYZ

The data breach tracking service Constella Intelligence reports that an Ivan V. Neculiti registered multiple online accounts under the email address dfyz_bk@bk.ru. Cyber intelligence firm Intel 471 shows this email address is tied to the username “dfyz” on more than a half-dozen Russian language cybercrime forums since 2008. The user dfyz on Searchengines[.]ru in 2008 asked other forum members to review war.md, and said they were part of the MercenarieS TeaM.

Back then, dfyz was selling “bulletproof servers for any purpose,” meaning the hosting company would willfully ignore abuse complaints or police inquiries about the activity of its customers.

DomainTools reports there are at least 33 domain names registered to dfyz_bk@bk.ru. Several of these domains have Ivan Neculiti in their registration records, including tracker-free[.]cn, which was registered to an Ivan Neculiti at dfyz_bk@bk.ru and referenced the MercenarieS TeaM in its original registration records.

Dfyz also used the nickname DonChicho, who likewise sold bulletproof hosting services and access to hacked Internet servers. In 2014, a prominent member of the Russian language cybercrime community Antichat filed a complaint against DonChicho, saying this user scammed them and had used the email address dfyz_bk@bk.ru.

The complaint said DonChicho registered on Antichat from the Transnistria Internet address 84.234.55[.]29. Searching this address in Constella reveals it has been used to register just five accounts online that have been created over the years, including one at ask.ru, where the user registered with the email address neculitzy1@yandex.ru. Constella also returns for that email address a user by the name “Ivan” at memoraleak.com and 000webhost.com.

Constella finds that the password most frequently used by the email address dfyz_bk@bk.ru was “filecast,” and that there are more than 90 email addresses associated with this password. Among them are roughly two dozen addresses with the name “Neculiti” in them, as well as the address support@donservers[.]ru.

Intel 471 says DonChicho posted to several Russian cybercrime forums that support@donservers[.]ru was his address, and that he logged into cybercrime forums almost exclusively from Internet addresses in Tiraspol, the capital of Transnistria. A review of DonChicho’s posts shows this person was banned from several forums in 2014 for scamming other users.

Cached copies of DonChicho’s vanity domain (donchicho[.]ru) show that in 2009 he was a spammer who peddled knockoff prescription drugs via Rx-Promotion, once one of the largest pharmacy spam moneymaking programs for Russian-speaking affiliates.

Mr. Neculiti told KrebsOnSecurity he has never used the nickname DonChicho.

“I may assure you that I have no relation to DonChicho nor to his bulletproof servers,” he said.

Below is a mind map that shows the connections between the accounts mentioned above.

Earlier this year, NoName began massively hitting government and industry websites in Moldova. A new report from Arbor Networks says the attacks began around March 6, when NoName alleged the government of Moldova was “craving for Russophobia.”

“Since early March, more than 50 websites have been targeted, according to posted ‘proof’ by the groups involved in attacking the country,” Arbor’s ASERT Team wrote. “While NoName seemingly initiated the ramp of attacks, a host of other DDoS hacktivists have joined the fray in claiming credit for attacks across more than 15 industries.”

CORRECTIV ACTION

The German independent news outlet Correctiv.org last week published a scathing investigative report on Stark Industries and MIRhosting, which notes that Ivan Neculiti operates his hosting companies with the help of his brother, Yuri.

The report points out that Stark Industries continues to host a Russian disinformation news outlet called “Recent Reliable News” (RRN) that was sanctioned by the European Union in 2023 for spreading links to propaganda blogs and fake European media and government websites.

“The website was not running on computers in Moscow or St. Petersburg until recently, but in the middle of the EU, in the Netherlands, on the computers of the Neculiti brothers,” Correctiv reporters wrote.

“After a request from this editorial team, a well-known service was installed that hides the actual web host,” the report continues. “Ivan Neculiti announced that he had blocked the associated access and server following internal investigations. “We very much regret that we are only now finding out that one of our customers is a sanctioned portal,” said the company boss. However, RRN is still accessible via its servers.”

Correctiv also points to a January 2023 report from the Ukrainian government, which found servers from Stark Industries Solutions were used as part of a cyber attack on the Ukrainian news agency “Ukrinform”. Correctiv notes the notorious hacker group Sandworm — an advanced persistent threat (APT) group operated by a cyberwarfare unit of Russia’s military intelligence service — was identified by Ukrainian government authorities as responsible for that attack.

PEACE HOSTING?

Public records indicate MIRhosting is based in The Netherlands and is operated by 37-year old Andrey Nesterenko, whose personal website says he is an accomplished concert pianist who began performing publicly at a young age.

DomainTools says mirhosting[.]com is registered to Mr. Nesterenko and to Innovation IT Solutions Corp, which lists addresses in London and in Nesterenko’s stated hometown of Nizhny Novgorod, Russia.

This is interesting because according to the book Inside Cyber Warfare by Jeffrey Carr, Innovation IT Solutions Corp. was responsible for hosting StopGeorgia[.]ru, a hacktivist website for organizing cyberattacks against Georgia that appeared at the same time Russian forces invaded the former Soviet nation in 2008. That conflict was thought to be the first war ever fought in which a notable cyberattack and an actual military engagement happened simultaneously.

Responding to questions from KrebsOnSecurity, Mr. Nesterenko said he couldn’t say whether his network had ever hosted the StopGeorgia website back in 2008 because his company didn’t keep records going back that far. But he said Stark Industries Solutions is indeed one of MIRhsoting’s colocation customers.

“Our relationship is purely provider-customer,” Nesterenko said. “They also utilize multiple providers and data centers globally, so connecting them directly to MIRhosting overlooks their broader network.”

“We take any report of malicious activity seriously and are always open to information that can help us identify and prevent misuse of our infrastructure, whether involving Stark Industries or any other customer,” Nesterenko continued. “In cases where our services are exploited for malicious purposes, we collaborate fully with Dutch cyber police and other relevant authorities to investigate and take appropriate measures. However, we have yet to receive any actionable information beyond the article itself, which has not provided us with sufficient detail to identify or block malicious actors.”

In December 2022, security firm Recorded Future profiled the phishing and credential harvesting infrastructure used for Russia-aligned espionage operations by a group dubbed Blue Charlie (aka TAG-53), which has targeted email accounts of nongovernmental organizations and think tanks, journalists, and government and defense officials.

Recorded Future found that virtually all the Blue Charlie domains existed in just ten different ISPs, with a significant concentration located in two networks, one of which was MIRhosting. Both Microsoft and the UK government assess that Blue Charlie is linked to the Russian threat activity groups variously known as Callisto Group, COLDRIVER, and SEABORGIUM.

Mr. Nesterenko took exception to a story on that report from The Record, which is owned by Recorded Future.

“We’ve discussed its contents with our customer, Stark Industries,” he said. “We understand that they have initiated legal proceedings against the website in question, as they firmly believe that the claims made are inaccurate.”

Recorded Future said they updated their story with comments from Mr. Neculiti, but that they stand by their reporting.

Mr. Nesterenko’s LinkedIn profile says he was previously the foreign region sales manager at Serverius-as, a hosting company in The Netherlands that remains in the same data center as MIRhosting.

In February, the Dutch police took 13 servers offline that were used by the infamous LockBit ransomware group, which had originally bragged on its darknet website that its home base was in The Netherlands. Sources tell KrebsOnSecurity the servers seized by the Dutch police were located in Serverius’ data center in Dronten, which is also shared by MIRhosting.

Serverius-as did not respond to requests for comment. Nesterenko said MIRhosting does use one of Serverius’s data centers for its operations in the Netherlands, alongside two other data centers, but that the recent incident involving the seizure of servers has no connection to MIRhosting.

“We are legally prohibited by Dutch law and police regulations from sharing information with third parties regarding any communications we may have had,” he said.

A February 2024 report from security firm ESET found Serverius-as systems were involved in a series of targeted phishing attacks by Russia-aligned groups against Ukrainian entities throughout 2023. ESET observed that after the spearphishing domains were no longer active, they were converted to promoting rogue Internet pharmacy websites.

PEERING INTO THE VOID

A review of the Internet address ranges recently added to the network operated by Stark Industries Solutions offers some insight into its customer base, usage, and maybe even true origins. Here is a snapshot (PDF) of all Internet address ranges announced by Stark Industries so far in the month of May 2024 (this information was graciously collated by the network observability platform Kentik.com).

Those records indicate that the largest portion of the IP space used by Stark is in The Netherlands, followed by Germany and the United States. Stark says it is connected to roughly 4,600 Internet addresses that currently list their ownership as Comcast Cable Communications.

A review of those address ranges at spur.us shows all of them are connected to an entity called Proxyline, which is a sprawling proxy service based in Russia that currently says it has more than 1.6 million proxies globally that are available for rent.

Reached for comment, Comcast said the Internet address ranges never did belong to Comcast, so it is likely that Stark has been fudging the real location of its routing announcements in some cases.

Stark reports that it has more than 67,000 Internet addresses at Santa Clara, Calif.-based EGIhosting. Spur says the Stark addresses involving EGIhosting all map to Proxyline as well. EGIhosting did not respond to requests for comment.

EGIhosting manages Internet addresses for the Cyprus-based hosting firm ITHOSTLINE LTD (aka HOSTLINE-LTD), which is represented throughout Stark’s announced Internet ranges. Stark says it has more than 21,000 Internet addresses with HOSTLINE. Spur.us finds Proxyline addresses are especially concentrated in the Stark ranges labeled ITHOSTLINE LTD, HOSTLINE-LTD, and Proline IT.

Stark’s network list includes approximately 21,000 Internet addresses at Hockessin, De. based DediPath, which abruptly ceased operations without warning in August 2023. According to a phishing report released last year by Interisle Consulting, DediPath was the fourth most common source of phishing attacks in the year ending Oct. 2022. Spur.us likewise finds that virtually all of the Stark address ranges marked “DediPath LLC” are tied to Proxyline.

A large number of the Internet address ranges announced by Stark in May originate in India, and the names that are self-assigned to many of these networks indicate they were previously used to send large volumes of spam for herbal medicinal products, with names like HerbalFarm, AdsChrome, Nutravo, Herbzoot and Herbalve.

The anti-spam organization SpamHaus reports that many of the Indian IP address ranges are associated with known “snowshoe spam,” a form of abuse that involves mass email campaigns spread across several domains and IP addresses to weaken reputation metrics and avoid spam filters.

It’s not clear how much of Stark’s network address space traces its origins to Russia, but big chunks of it recently belonged to some of the oldest entities on the Russian Internet (a.k.a. “Runet”).

For example, many Stark address ranges were most recently assigned to a Russian government entity whose full name is the “Federal State Autonomous Educational Establishment of Additional Professional Education Center of Realization of State Educational Policy and Informational Technologies.”

A review of Internet address ranges adjacent to this entity reveals a long list of Russian government organizations that are part of the Federal Guard Service of the Russian Federation. Wikipedia says the Federal Guard Service is a Russian federal government agency concerned with tasks related to protection of several high-ranking state officials, including the President of Russia, as well as certain federal properties. The agency traces its origins to the USSR’s Ninth Directorate of the KGB, and later the presidential security service.

Stark recently announced the address range 213.159.64.0/20 from April 27 to May 1, and this range was previously assigned to an ancient ISP in St. Petersburg, RU called the Computer Technologies Institute Ltd.

According to a post on the Russian language webmaster forum searchengines[.]ru, the domain for Computer Technologies Institute — ctinet[.]ru — is the seventh-oldest domain in the entire history of the Runet.

Curiously, Stark also lists large tracts of Internet addresses (close to 48,000 in total) assigned to a small ISP in Kharkiv, Ukraine called NetAssist. Reached via email, the CEO of NetAssist Max Tulyev confirmed his company provides a number of services to PQ Hosting.

“We colocate their equipment in Warsaw, Madrid, Sofia and Thessaloniki, provide them IP transit and IPv4 addresses,” Tulyev said. “For their size, we receive relatively low number of complains to their networks. I never seen anything about their pro-Russian activity or support of Russian hackers. It is very interesting for me to see proofs of your accusations.”

Spur.us mapped the entire infrastructure of Proxyline, and found more than one million proxies across multiple providers, but by far the biggest concentration was at Stark Industries Solutions. The full list of Proxyline address ranges (.CSV) shows two other ISPs appear repeatedly throughout the list. One is Kharkiv, Ukraine based ITL LLC, also known as Information Technology Laboratories Group, and Integrated Technologies Laboratory.

The second is a related hosting company in Miami, called Green Floid LLC. Green Floid featured in a 2017 scoop by CNN, which profiled the company’s owner and quizzed him about Russian troll farms using proxy networks on Green Floid and its parent firm ITL to mask disinformation efforts tied to the Kremlin’s Internet Research Agency (IRA). At the time, the IRA was using Facebook and other social media networks to spread videos showing police brutality against African Americans in an effort to encourage protests across the United States.

Doug Madory, director of Internet analysis at Kentik, was able to see at a high level the top sources and destinations for traffic traversing Stark’s network.

“Based on our aggregate NetFlow, we see Iran as the top destination (35.1%) for traffic emanating from Stark (AS44477),” Madory said. “Specifically, the top destination is MTN Irancell, while the top source is Facebook. This data supports the theory that AS44477 houses proxy services as Facebook is blocked in Iran.”

On April 30, the security firm Malwarebytes explored an extensive malware operation that targets corporate Internet users with malicious ads. Among the sites used as lures in that campaign were fake Wall Street Journal and CNN websites that told visitors they were required to install a WSJ or CNN-branded browser extension (malware). Malwarebytes found a domain name central to that operation was hosted at Internet addresses owned by Stark Industries.

Patch Tuesday, May 2024 Edition

Microsoft today released updates to fix more than 60 security holes in Windows computers and supported software, including two “zero-day” vulnerabilities in Windows that are already being exploited in active attacks. There are also important security patches available for macOS and Adobe users, and for the Chrome Web browser, which just patched its own zero-day flaw.

First, the zero-days. CVE-2024-30051 is an “elevation of privilege” bug in a core Windows library. Satnam Narang at Tenable said this flaw is being used as part of post-compromise activity to elevate privileges as a local attacker.

“CVE-2024-30051 is used to gain initial access into a target environment and requires the use of social engineering tactics via email, social media or instant messaging to convince a target to open a specially crafted document file,” Narang said. “Once exploited, the attacker can bypass OLE mitigations in Microsoft 365 and Microsoft Office, which are security features designed to protect end users from malicious files.”

Kaspersky Lab, one of two companies credited with reporting exploitation of CVE-2024-30051 to Microsoft, has published a fascinating writeup on how they discovered the exploit in a file shared with Virustotal.com.

Kaspersky said it has since seen the exploit used together with QakBot and other malware. Emerging in 2007 as a banking trojan, QakBot (a.k.a. Qbot and Pinkslipbot) has morphed into an advanced malware strain now used by multiple cybercriminal groups to prepare newly compromised networks for ransomware infestations.

CVE-2024-30040 is a security feature bypass in MSHTML, a component that is deeply tied to the default Web browser on Windows systems. Microsoft’s advisory on this flaw is fairly sparse, but Kevin Breen from Immersive Labs said this vulnerability also affects Office 365 and Microsoft Office applications.

“Very little information is provided and the short description is painfully obtuse,” Breen said of Microsoft’s advisory on CVE-2024-30040.

The only vulnerability fixed this month that earned Microsoft’s most-dire “critical” rating is CVE-2024-30044, a flaw in Sharepoint that Microsoft said is likely to be exploited. Tenable’s Narang notes that exploitation of this bug requires an attacker to be authenticated to a vulnerable SharePoint Server with Site Owner permissions (or higher) first and to take additional steps in order to exploit this flaw, which makes this flaw less likely to be widely exploited as most attackers follow the path of least resistance.

Five days ago, Google released a security update for Chrome that fixes a zero-day in the popular browser. Chrome usually auto-downloads any available updates, but it still may require a complete restart of the browser to install them. If you use Chrome and see a “Relaunch to update” message in the upper right corner of the browser, it’s time to restart.

Apple has just shipped macOS Sonoma 14.5 update, which includes nearly two dozen security patches. To ensure your Mac is up-to-date, go to System Settings, General tab, then Software Update and follow any prompts.

Finally, Adobe has critical security patches available for a range of products, including Acrobat, Reader, Illustrator, Adobe Substance 3D Painter, Adobe Aero, Adobe Animate and Adobe Framemaker.

Regardless of whether you use a Mac or Windows system (or something else), it’s always a good idea to backup your data and or system before applying any security updates. For a closer look at the individual fixes released by Microsoft today, check out the complete list over at the SANS Internet Storm Center. Anyone in charge of maintaining Windows systems in an enterprise environment should keep an eye on askwoody.com, which usually has the scoop on any wonky Windows patches.

Update, May 15, 8:28 a.m.: Corrected misattribution of CVE-2024-30051.

Why Your VPN May Not Be As Secure As It Claims

Virtual private networking (VPN) companies market their services as a way to prevent anyone from snooping on your Internet usage. But new research suggests this is a dangerous assumption when connecting to a VPN via an untrusted network, because attackers on the same network could force a target’s traffic off of the protection provided by their VPN without triggering any alerts to the user.

When a device initially tries to connect to a network, it broadcasts a message to the entire local network stating that it is requesting an Internet address. Normally, the only system on the network that notices this request and replies is the router responsible for managing the network to which the user is trying to connect.

The machine on a network responsible for fielding these requests is called a Dynamic Host Configuration Protocol (DHCP) server, which will issue time-based leases for IP addresses. The DHCP server also takes care of setting a specific local address — known as an Internet gateway — that all connecting systems will use as a primary route to the Web.

VPNs work by creating a virtual network interface that serves as an encrypted tunnel for communications. But researchers at Leviathan Security say they’ve discovered it’s possible to abuse an obscure feature built into the DHCP standard so that other users on the local network are forced to connect to a rogue DHCP server.

“Our technique is to run a DHCP server on the same network as a targeted VPN user and to also set our DHCP configuration to use itself as a gateway,” Leviathan researchers Lizzie Moratti and Dani Cronce wrote. “When the traffic hits our gateway, we use traffic forwarding rules on the DHCP server to pass traffic through to a legitimate gateway while we snoop on it.”

The feature being abused here is known as DHCP option 121, and it allows a DHCP server to set a route on the VPN user’s system that is more specific than those used by most VPNs. Abusing this option, Leviathan found, effectively gives an attacker on the local network the ability to set up routing rules that have a higher priority than the routes for the virtual network interface that the target’s VPN creates.

“Pushing a route also means that the network traffic will be sent over the same interface as the DHCP server instead of the virtual network interface,” the Leviathan researchers said. “This is intended functionality that isn’t clearly stated in the RFC [standard]. Therefore, for the routes we push, it is never encrypted by the VPN’s virtual interface but instead transmitted by the network interface that is talking to the DHCP server. As an attacker, we can select which IP addresses go over the tunnel and which addresses go over the network interface talking to our DHCP server.”

Leviathan found they could force VPNs on the local network that already had a connection to arbitrarily request a new one. In this well-documented tactic, known as a DHCP starvation attack, an attacker floods the DHCP server with requests that consume all available IP addresses that can be allocated. Once the network’s legitimate DHCP server is completely tied up, the attacker can then have their rogue DHCP server respond to all pending requests.

“This technique can also be used against an already established VPN connection once the VPN user’s host needs to renew a lease from our DHCP server,” the researchers wrote. “We can artificially create that scenario by setting a short lease time in the DHCP lease, so the user updates their routing table more frequently. In addition, the VPN control channel is still intact because it already uses the physical interface for its communication. In our testing, the VPN always continued to report as connected, and the kill switch was never engaged to drop our VPN connection.”

The researchers say their methods could be used by an attacker who compromises a DHCP server or wireless access point, or by a rogue network administrator who owns the infrastructure themselves and maliciously configures it. Alternatively, an attacker could set up an “evil twin” wireless hotspot that mimics the signal broadcast by a legitimate provider.

ANALYSIS

Bill Woodcock is executive director at Packet Clearing House, a nonprofit based in San Francisco. Woodcock said Option 121 has been included in the DHCP standard since 2002, which means the attack described by Leviathan has technically been possible for the last 22 years.

“They’re realizing now that this can be used to circumvent a VPN in a way that’s really problematic, and they’re right,” Woodcock said.

Woodcock said anyone who might be a target of spear phishing attacks should be very concerned about using VPNs on an untrusted network.

“Anyone who is in a position of authority or maybe even someone who is just a high net worth individual, those are all very reasonable targets of this attack,” he said. “If I were trying to do an attack against someone at a relatively high security company and I knew where they typically get their coffee or sandwich at twice a week, this is a very effective tool in that toolbox. I’d be a little surprised if it wasn’t already being exploited in that way, because again this isn’t rocket science. It’s just thinking a little outside the box.”

Successfully executing this attack on a network likely would not allow an attacker to see all of a target’s traffic or browsing activity. That’s because for the vast majority of the websites visited by the target, the content is encrypted (the site’s address begins with https://). However, an attacker would still be able to see the metadata — such as the source and destination addresses — of any traffic flowing by.

KrebsOnSecurity shared Leviathan’s research with John Kristoff, founder of dataplane.org and a PhD candidate in computer science at the University of Illinois Chicago. Kristoff said practically all user-edge network gear, including WiFi deployments, support some form of rogue DHCP server detection and mitigation, but that it’s unclear how widely deployed those protections are in real-world environments.

“However, and I think this is a key point to emphasize, an untrusted network is an untrusted network, which is why you’re usually employing the VPN in the first place,” Kristoff said. “If [the] local network is inherently hostile and has no qualms about operating a rogue DHCP server, then this is a sneaky technique that could be used to de-cloak some traffic – and if done carefully, I’m sure a user might never notice.”

MITIGATIONS

According to Leviathan, there are several ways to minimize the threat from rogue DHCP servers on an unsecured network. One is using a device powered by the Android operating system, which apparently ignores DHCP option 121.

Relying on a temporary wireless hotspot controlled by a cellular device you own also effectively blocks this attack.

“They create a password-locked LAN with automatic network address translation,” the researchers wrote of cellular hot-spots. “Because this network is completely controlled by the cellular device and requires a password, an attacker should not have local network access.”

Leviathan’s Moratti said another mitigation is to run your VPN from inside of a virtual machine (VM) — like Parallels, VMware or VirtualBox. VPNs run inside of a VM are not vulnerable to this attack, Moratti said, provided they are not run in “bridged mode,” which causes the VM to replicate another node on the network.

In addition, a technology called “deep packet inspection” can be used to deny all in- and outbound traffic from the physical interface except for the DHCP and the VPN server. However, Leviathan says this approach opens up a potential “side channel” attack that could be used to determine the destination of traffic.

“This could be theoretically done by performing traffic analysis on the volume a target user sends when the attacker’s routes are installed compared to the baseline,” they wrote. “In addition, this selective denial-of-service is unique as it could be used to censor specific resources that an attacker doesn’t want a target user to connect to even while they are using the VPN.”

Moratti said Leviathan’s research shows that many VPN providers are currently making promises to their customers that their technology can’t keep.

“VPNs weren’t designed to keep you more secure on your local network, but to keep your traffic more secure on the Internet,” Moratti said. “When you start making assurances that your product protects people from seeing your traffic, there’s an assurance or promise that can’t be met.”

A copy of Leviathan’s research, along with code intended to allow others to duplicate their findings in a lab environment, is available here.

April’s Patch Tuesday Brings Record Number of Fixes

If only Patch Tuesdays came around infrequently — like total solar eclipse rare — instead of just creeping up on us each month like The Man in the Moon. Although to be fair, it would be tough for Microsoft to eclipse the number of vulnerabilities fixed in this month’s patch batch — a record 147 flaws in Windows and related software.

Yes, you read that right. Microsoft today released updates to address 147 security holes in Windows, Office, Azure, .NET Framework, Visual Studio, SQL Server, DNS Server, Windows Defender, Bitlocker, and Windows Secure Boot.

“This is the largest release from Microsoft this year and the largest since at least 2017,” said Dustin Childs, from Trend Micro’s Zero Day Initiative (ZDI). “As far as I can tell, it’s the largest Patch Tuesday release from Microsoft of all time.”

Tempering the sheer volume of this month’s patches is the middling severity of many of the bugs. Only three of April’s vulnerabilities earned Microsoft’s most-dire “critical” rating, meaning they can be abused by malware or malcontents to take remote control over unpatched systems with no help from users.

Most of the flaws that Microsoft deems “more likely to be exploited” this month are marked as “important,” which usually involve bugs that require a bit more user interaction (social engineering) but which nevertheless can result in system security bypass, compromise, and the theft of critical assets.

Ben McCarthy, lead cyber security engineer at Immersive Labs called attention to CVE-2024-20670, an Outlook for Windows spoofing vulnerability described as being easy to exploit. It involves convincing a user to click on a malicious link in an email, which can then steal the user’s password hash and authenticate as the user in another Microsoft service.

Another interesting bug McCarthy pointed to is CVE-2024-29063, which involves hard-coded credentials in Azure’s search backend infrastructure that could be gleaned by taking advantage of Azure AI search.

“This along with many other AI attacks in recent news shows a potential new attack surface that we are just learning how to mitigate against,” McCarthy said. “Microsoft has updated their backend and notified any customers who have been affected by the credential leakage.”

CVE-2024-29988 is a weakness that allows attackers to bypass Windows SmartScreen, a technology Microsoft designed to provide additional protections for end users against phishing and malware attacks. Childs said one of ZDI’s researchers found this vulnerability being exploited in the wild, although Microsoft doesn’t currently list CVE-2024-29988 as being exploited.

“I would treat this as in the wild until Microsoft clarifies,” Childs said. “The bug itself acts much like CVE-2024-21412 – a [zero-day threat from February] that bypassed the Mark of the Web feature and allows malware to execute on a target system. Threat actors are sending exploits in a zipped file to evade EDR/NDR detection and then using this bug (and others) to bypass Mark of the Web.”

Update, 7:46 p.m. ET: A previous version of this story said there were no zero-day vulnerabilities fixed this month. BleepingComputer reports that Microsoft has since confirmed that there are actually two zero-days. One is the flaw Childs just mentioned (CVE-2024-21412), and the other is CVE-2024-26234, described as a “proxy driver spoofing” weakness.

Satnam Narang at Tenable notes that this month’s release includes fixes for two dozen flaws in Windows Secure Boot, the majority of which are considered “Exploitation Less Likely” according to Microsoft.

“However, the last time Microsoft patched a flaw in Windows Secure Boot in May 2023 had a notable impact as it was exploited in the wild and linked to the BlackLotus UEFI bootkit, which was sold on dark web forums for $5,000,” Narang said. “BlackLotus can bypass functionality called secure boot, which is designed to block malware from being able to load when booting up. While none of these Secure Boot vulnerabilities addressed this month were exploited in the wild, they serve as a reminder that flaws in Secure Boot persist, and we could see more malicious activity related to Secure Boot in the future.”

For links to individual security advisories indexed by severity, check out ZDI’s blog and the Patch Tuesday post from the SANS Internet Storm Center. Please consider backing up your data or your drive before updating, and drop a note in the comments here if you experience any issues applying these fixes.

Adobe today released nine patches tackling at least two dozen vulnerabilities in a range of software products, including Adobe After Effects, Photoshop, Commerce, InDesign, Experience Manager, Media Encoder, Bridge, Illustrator, and Adobe Animate.

KrebsOnSecurity needs to correct the record on a point mentioned at the end of March’s “Fat Patch Tuesday” post, which looked at new AI capabilities built into Adobe Acrobat that are turned on by default. Adobe has since clarified that its apps won’t use AI to auto-scan your documents, as the original language in its FAQ suggested.

“In practice, no document scanning or analysis occurs unless a user actively engages with the AI features by agreeing to the terms, opening a document, and selecting the AI Assistant or generative summary buttons for that specific document,” Adobe said earlier this month.

Feds Seize LockBit Ransomware Websites, Offer Decryption Tools, Troll Affiliates

U.S. and U.K. authorities have seized the darknet websites run by LockBit, a prolific and destructive ransomware group that has claimed more than 2,000 victims worldwide and extorted over $120 million in payments. Instead of listing data stolen from ransomware victims who didn’t pay, LockBit’s victim shaming website now offers free recovery tools, as well as news about arrests and criminal charges involving LockBit affiliates.

Dubbed “Operation Cronos,” the law enforcement action involved the seizure of nearly three-dozen servers; the arrest of two alleged LockBit members; the unsealing of two indictments; the release of a free LockBit decryption tool; and the freezing of more than 200 cryptocurrency accounts thought to be tied to the gang’s activities.

LockBit members have executed attacks against thousands of victims in the United States and around the world, according to the U.S. Department of Justice (DOJ). First surfacing in September 2019, the gang is estimated to have made hundreds of millions of U.S. dollars in ransom demands, and extorted over $120 million in ransom payments.

LockBit operated as a ransomware-as-a-service group, wherein the ransomware gang takes care of everything from the bulletproof hosting and domains to the development and maintenance of the malware. Meanwhile, affiliates are solely responsible for finding new victims, and can reap 60 to 80 percent of any ransom amount ultimately paid to the group.

A statement on Operation Cronos from the European police agency Europol said the months-long infiltration resulted in the compromise of LockBit’s primary platform and other critical infrastructure, including the takedown of 34 servers in the Netherlands, Germany, Finland, France, Switzerland, Australia, the United States and the United Kingdom. Europol said two suspected LockBit actors were arrested in Poland and Ukraine, but no further information has been released about those detained.

The DOJ today unsealed indictments against two Russian men alleged to be active members of LockBit. The government says Russian national Artur Sungatov used LockBit ransomware against victims in manufacturing, logistics, insurance and other companies throughout the United States.

Ivan Gennadievich Kondratyev, a.k.a. “Bassterlord,” allegedly deployed LockBit against targets in the United States, Singapore, Taiwan, and Lebanon. Kondratyev is also charged (PDF) with three criminal counts arising from his alleged use of the Sodinokibi (aka “REvil“) ransomware variant to encrypt data, exfiltrate victim information, and extort a ransom payment from a corporate victim based in Alameda County, California.

With the indictments of Sungatov and Kondratyev, a total of five LockBit affiliates now have been officially charged. In May 2023, U.S. authorities unsealed indictments against two alleged LockBit affiliates, Mikhail “Wazawaka” Matveev and Mikhail Vasiliev.

Vasiliev, 35, of Bradford, Ontario, Canada, is in custody in Canada awaiting extradition to the United States (the complaint against Vasiliev is at this PDF). Matveev remains at large, presumably still in Russia. In January 2022, KrebsOnSecurity published Who is the Network Access Broker ‘Wazawaka,’ which followed clues from Wazawaka’s many pseudonyms and contact details on the Russian-language cybercrime forums back to a 31-year-old Mikhail Matveev from Abaza, RU.

In June 2023, Russian national Ruslan Magomedovich Astamirov was charged in New Jersey for his participation in the LockBit conspiracy, including the deployment of LockBit against victims in Florida, Japan, France, and Kenya. Astamirov is currently in custody in the United States awaiting trial.

LockBit was known to have recruited affiliates that worked with multiple ransomware groups simultaneously, and it’s unclear what impact this takedown may have on competing ransomware affiliate operations. The security firm ProDaft said on Twitter/X that the infiltration of LockBit by investigators provided “in-depth visibility into each affiliate’s structures, including ties with other notorious groups such as FIN7, Wizard Spider, and EvilCorp.”

In a lengthy thread about the LockBit takedown on the Russian-language cybercrime forum XSS, one of the gang’s leaders said the FBI and the U.K.’s National Crime Agency (NCA) had infiltrated its servers using a known vulnerability in PHP, a scripting language that is widely used in Web development.

Several denizens of XSS wondered aloud why the PHP flaw was not flagged by LockBit’s vaunted “Bug Bounty” program, which promised a financial reward to affiliates who could find and quietly report any security vulnerabilities threatening to undermine LockBit’s online infrastructure.

This prompted several XSS members to start posting memes taunting the group about the security failure.

“Does it mean that the FBI provided a pentesting service to the affiliate program?,” one denizen quipped. “Or did they decide to take part in the bug bounty program? :):)”

Federal investigators also appear to be trolling LockBit members with their seizure notices. LockBit’s data leak site previously featured a countdown timer for each victim organization listed, indicating the time remaining for the victim to pay a ransom demand before their stolen files would be published online. Now, the top entry on the shaming site is a countdown timer until the public doxing of “LockBitSupp,” the unofficial spokesperson or figurehead for the LockBit gang.

“Who is LockbitSupp?” the teaser reads. “The $10m question.”

In January 2024, LockBitSupp told XSS forum members he was disappointed the FBI hadn’t offered a reward for his doxing and/or arrest, and that in response he was placing a bounty on his own head — offering $10 million to anyone who could discover his real name.

“My god, who needs me?,” LockBitSupp wrote on Jan. 22, 2024. “There is not even a reward out for me on the FBI website. By the way, I want to use this chance to increase the reward amount for a person who can tell me my full name from USD 1 million to USD 10 million. The person who will find out my name, tell it to me and explain how they were able to find it out will get USD 10 million. Please take note that when looking for criminals, the FBI uses unclear wording offering a reward of UP TO USD 10 million; this means that the FBI can pay you USD 100, because technically, it’s an amount UP TO 10 million. On the other hand, I am willing to pay USD 10 million, no more and no less.”

Mark Stockley, cybersecurity evangelist at the security firm Malwarebytes, said the NCA is obviously trolling the LockBit group and LockBitSupp.

“I don’t think this is an accident—this is how ransomware groups talk to each other,” Stockley said. “This is law enforcement taking the time to enjoy its moment, and humiliate LockBit in its own vernacular, presumably so it loses face.”

In a press conference today, the FBI said Operation Cronos included investigative assistance from the Gendarmerie-C3N in France; the State Criminal Police Office L-K-A and Federal Criminal Police Office in Germany; Fedpol and Zurich Cantonal Police in Switzerland; the National Police Agency in Japan; the Australian Federal Police; the Swedish Police Authority; the National Bureau of Investigation in Finland; the Royal Canadian Mounted Police; and the National Police in the Netherlands.

The Justice Department said victims targeted by LockBit should contact the FBI at https://lockbitvictims.ic3.gov/ to determine whether affected systems can be successfully decrypted. In addition, the Japanese Police, supported by Europol, have released a recovery tool designed to recover files encrypted by the LockBit 3.0 Black Ransomware.

ICANN Launches Service to Help With WHOIS Lookups

More than five years after domain name registrars started redacting personal data from all public domain registration records, the non-profit organization overseeing the domain industry has introduced a centralized online service designed to make it easier for researchers, law enforcement and others to request the information directly from registrars.

In May 2018, the Internet Corporation for Assigned Names and Numbers (ICANN) — the nonprofit entity that manages the global domain name system — instructed all registrars to redact the customer’s name, address, phone number and email from WHOIS, the system for querying databases that store the registered users of domain names and blocks of Internet address ranges.

ICANN made the policy change in response to the General Data Protection Regulation (GDPR), a law enacted by the European Parliament that requires companies to gain affirmative consent for any personal information they collect on people within the European Union. In the meantime, registrars were to continue collecting the data but not publish it, and ICANN promised it would develop a system that facilitates access to this information.

At the end of November 2023, ICANN launched the Registration Data Request Service (RDRS), which is designed as a one-stop shop to submit registration data requests to participating registrars. This video from ICANN walks through how the system works.

Accredited registrars don’t have to participate, but ICANN is asking all registrars to join and says participants can opt out or stop using it at any time. ICANN contends that the use of a standardized request form makes it easier for the correct information and supporting documents to be provided to evaluate a request.

ICANN says the RDRS doesn’t guarantee access to requested registration data, and that all communication and data disclosure between the registrars and requestors takes place outside of the system. The service can’t be used to request WHOIS data tied to country-code top level domains (CCTLDs), such as those ending in .de (Germany) or .nz (New Zealand), for example.

As Catalin Cimpanu writes for Risky Business News, currently investigators can file legal requests or abuse reports with each individual registrar, but the idea behind the RDRS is to create a place where requests from “verified” parties can be honored faster and with a higher degree of trust.

The registrar community generally views public WHOIS data as a nuisance issue for their domain customers and an unwelcome cost-center. Privacy advocates maintain that cybercriminals don’t provide their real information in registration records anyway, and that requiring WHOIS data to be public simply causes domain registrants to be pestered by spammers, scammers and stalkers.

Meanwhile, security experts argue that even in cases where online abusers provide intentionally misleading or false information in WHOIS records, that information is still extremely useful in mapping the extent of their malware, phishing and scamming operations. What’s more, the overwhelming majority of phishing is performed with the help of compromised domains, and the primary method for cleaning up those compromises is using WHOIS data to contact the victim and/or their hosting provider.

Anyone looking for copious examples of both need only to search this Web site for the term “WHOIS,” which yields dozens of stories and investigations that simply would not have been possible without the data available in the global WHOIS records.

KrebsOnSecurity remains doubtful that participating registrars will be any more likely to share WHOIS data with researchers just because the request comes through ICANN. But I look forward to being wrong on this one, and will certainly mention it in my reporting if the RDRS proves useful.

Regardless of whether the RDRS succeeds or fails, there is another European law that takes effect in 2024 which is likely to place additional pressure on registrars to respond to legitimate WHOIS data requests. The new Network and Information Security Directive (NIS2), which EU member states have until October 2024 to implement, requires registrars to keep much more accurate WHOIS records, and to respond within as little as 24 hours to WHOIS data requests tied everything from phishing, malware and spam to copyright and brand enforcement.

.US Harbors Prolific Malicious Link Shortening Service

The top-level domain for the United States — .US — is home to thousands of newly-registered domains tied to a malicious link shortening service that facilitates malware and phishing scams, new research suggests. The findings come close on the heels of a report that identified .US domains as among the most prevalent in phishing attacks over the past year.

Researchers at Infoblox say they’ve been tracking what appears to be a three-year-old link shortening service that is catering to phishers and malware purveyors. Infoblox found the domains involved are typically three to seven characters long, and hosted on bulletproof hosting providers that charge a premium to ignore any abuse or legal complaints. The short domains don’t host any content themselves, but are used to obfuscate the real address of landing pages that try to phish users or install malware.

Infoblox says it’s unclear how the phishing and malware landing pages tied to this service are being initially promoted, although they suspect it is mainly through scams targeting people on their phones via SMS. A new report says the company mapped the contours of this link shortening service thanks in part to pseudo-random patterns in the short domains, which all appear on the surface to be a meaningless jumble of letters and numbers.

“This came to our attention because we have systems that detect registrations that use domain name generation algorithms,” said Renee Burton, head of threat intelligence at Infoblox. “We have not found any legitimate content served through their shorteners.”

Infoblox determined that until May 2023, domains ending in .info accounted for the bulk of new registrations tied to the malicious link shortening service, which Infoblox has dubbed “Prolific Puma.” Since then, they found that whoever is responsible for running the service has used .US for approximately 55 percent of the total domains created, with several dozen new malicious .US domains registered daily.

.US is overseen by the National Telecommunications and Information Administration (NTIA), an executive branch agency of the U.S. Department of Commerce. But Uncle Sam has long outsourced the management of .US to various private companies, which have gradually allowed the United States’s top-level domain to devolve into a cesspool of phishing activity.

Or so concludes The Interisle Consulting Group, which gathers phishing data from multiple industry sources and publishes an annual report on the latest trends. As far back as 2018, Interisle found .US domains were the worst in the world for spam, botnet (attack infrastructure for DDOS etc.) and illicit or harmful content.

Interisle’s newest study examined six million phishing reports between May 1, 2022 and April 30, 2023, and identified approximately 30,000 .US phishing domains. Interisle found significant numbers of .US domains were registered to attack some of the United States’ most prominent companies, including Bank of America, Amazon, Apple, AT&T, Citi, Comcast, Microsoft, Meta, and Target. Others were used to impersonate or attack U.S. government agencies.

Under NTIA regulations, domain registrars processing .US domain registrations must take certain steps (PDF) to verify that those customers actually reside in the United States, or else own organizations based in the U.S. However, if one registers a .US domain through GoDaddy — the largest domain registrar and the current administrator of the .US contract — the way one “proves” their U.S. nexus is simply by choosing from one of three pre-selected affirmative responses.

In an age when most domain registrars are automatically redacting customer information from publicly accessible registration records to avoid running afoul of European privacy laws, .US has remained something of an outlier because its charter specifies that all registration records be made public. However, Infoblox said it found more than 2,000 malicious link shortener domains ending in .US registered since October 2023 through NameSilo that have somehow subverted the transparency requirements for the usTLD and converted to private registrations.

“Through our own experience with NameSilo, it is not possible to select private registration for domains in the usTLD through their interface,” Infoblox wrote. “And yet, it was done. Of the total domains with private records, over 99% were registered with NameSilo. At this time, we are not able to explain this behavior.”

NameSilo CEO Kristaps Ronka said the company actively responds to reports about abusive domains, but that it hasn’t seen any abuse reports related to Infoblox’s findings.

“We take down hundreds to thousands of domains, lots of them proactively to combat abuse,” Ronka said. “Our current abuse rate on abuseIQ for example is currently at 0%. AbuseIQ receives reports from countless sources and we are yet to see these ‘Puma’ abuse reports.”

Experts who track domains associated with malware and phishing say even phony information supplied at registration is useful in identifying potentially malicious or phishous domains before they can be used for abuse.

For example, when it was registered through NameSilo in July 2023, the domain 1ox[.]us — like thousands of others — listed its registrant as “Leila Puma” at a street address in Poland, and the email address blackpumaoct33@ukr.net. But according to DomainTools.com, on Oct. 1, 2023 those records were redacted and hidden by NameSilo.

Infoblox notes that the username portion of the email address appears to be a reference to the song October 33 by the Black Pumas, an Austin, Texas based psychedelic soul band. The Black Pumas aren’t exactly a household name, but they did recently have a popular Youtube video that featured a cover of the Kinks song “Strangers,” which included an emotional visual narrative about Ukrainians seeking refuge from the Russian invasion, titled “Ukraine Strangers.” Also, Leila Puma’s email address is at a Ukrainian email provider.

DomainTools shows that hundreds of other malicious domains tied to Prolific Puma previously were registered through NameCheap to a “Josef Bakhovsky” at a different street address in Poland. According to ancestry.com, the anglicized version of this surname — Bakovski — is the traditional name for someone from Bakowce, which is now known as Bakivtsi and is in Ukraine.

This possible Polish and/or Ukrainian connection may or may not tell us something about the “who” behind this link shortening service, but those details are useful for identifying and grouping these malicious short domains. However, even this meager visibility into .US registration data is now under threat.

The NTIA recently published a proposal that would allow registrars to redact all registrant data from WHOIS registration records for .US domains. A broad array of industry groups have filed comments opposing the proposed changes, saying they threaten to remove the last vestiges of accountability for a top-level domain that is already overrun with cybercrime activity.

Infoblox’s Burton says Prolific Puma is remarkable because they’ve been able to facilitate malicious activities for years while going largely unnoticed by the security industry.

“This exposes how persistent the criminal economy can be at a supply chain level,” Burton said. “We’re always looking at the end malware or phishing page, but what we’re finding here is that there’s this middle layer of DNS threat actors persisting for years without notice.”

Infoblox’s full report on Prolific Puma is here.

Why is .US Being Used to Phish So Many of Us?

Domain names ending in “.US” — the top-level domain for the United States — are among the most prevalent in phishing scams, new research shows. This is noteworthy because .US is overseen by the U.S. government, which is frequently the target of phishing domains ending in .US. Also, .US domains are only supposed to be available to U.S. citizens and to those who can demonstrate that they have a physical presence in the United States.

.US is the “country code top-level domain” or ccTLD of the United States. Most countries have their own ccTLDs: .MX for Mexico, for example, or .CA for Canada. But few other major countries in the world have anywhere near as many phishing domains each year as .US.

That’s according to The Interisle Consulting Group, which gathers phishing data from multiple industry sources and publishes an annual report on the latest trends. Interisle’s newest study examined six million phishing reports between May 1, 2022 and April 30, 2023, and found 30,000 .US phishing domains.

.US is overseen by the National Telecommunications and Information Administration (NTIA), an executive branch agency of the U.S. Department of Commerce. However, NTIA currently contracts out the management of the .US domain to GoDaddy, by far the world’s largest domain registrar.

Under NTIA regulations, the administrator of the .US registry must take certain steps to verify that their customers actually reside in the United States, or own organizations based in the U.S. But Interisle found that whatever GoDaddy was doing to manage that vetting process wasn’t working.

“The .US ‘nexus’ requirement theoretically limits registrations to parties with a national connection, but .US had very high numbers of phishing domains,” Interisle wrote. “This indicates a possible problem with the administration or application of the nexus requirements.”

Dean Marks is emeritus executive director for a group called the Coalition for Online Accountability, which has been critical of the NTIA’s stewardship of .US. Marks says virtually all European Union member state ccTLDs that enforce nexus restrictions also have massively lower levels of abuse due to their policies and oversight.

“Even very large ccTLDs, like .de for Germany — which has a far larger market share of domain name registrations than .US — have very low levels of abuse, including phishing and malware,” Marks told KrebsOnSecurity. “In my view, this situation with .US should not be acceptable to the U.S. government overall, nor to the US public.”

Marks said there are very few phishing domains ever registered in other ccTLDs that also restrict registrations to their citizens, such as .HU (Hungary), .NZ (New Zealand), and .FI (Finland), where a connection to the country, a proof of identity, or evidence of incorporation are required.

“Or .LK (Sri Lanka), where the acceptable use policy includes a ‘lock and suspend’ if domains are reported for suspicious activity,” Marks said. “These ccTLDs make a strong case for validating domain registrants in the interest of public safety.”

Sadly, .US has been a cesspool of phishing activity for many years. As far back as 2018, Interisle found .US domains were the worst in the world for spam, botnet (attack infrastructure for DDOS etc.) and illicit or harmful content. Back then, .US was being operated by a different contractor.

In response to questions from KrebsOnSecurity, GoDaddy said all .US registrants must certify that they meet the NTIA’s nexus requirements. But this appears to be little more than an affirmative response that is already pre-selected for all new registrants.

Attempting to register a .US domain through GoDaddy, for example, leads to a U.S. Registration Information page that auto-populates the nexus attestation field with the response, “I am a citizen of the US.” Other options include, “I am a permanent resident of the US,” and “My primary domicile is in the US.” It currently costs just $4.99 to obtain a .US domain through GoDaddy.

GoDaddy said it also conducts a scan of selected registration request information, and conducts “spot checks” on registrant information.

“We conduct regular reviews, per policy, of registration data within the Registry database to determine Nexus compliance with ongoing communications to registrars and registrants,” the company said in a written statement.

GoDaddy says it “is committed to supporting a safer online environment and proactively addressing this issue by assessing it against our own anti-abuse mitigation system.”

“We stand against DNS abuse in any form and maintain multiple systems and protocols to protect all the TLDs we operate,” the statement continued. “We will continue to work with registrars, cybersecurity firms and other stakeholders to make progress with this complex challenge.”

Interisle found significant numbers of .US domains were registered to attack some of the United States’ most prominent companies, including Bank of America, Amazon, Apple, AT&T, Citi, Comcast, Microsoft, Meta, and Target.

“Ironically, at least 109 of the .US domains in our data were used to attack the United States government, specifically the United States Postal Service and its customers,” Interisle wrote. “.US domains were also used to attack foreign government operations: six .US domains were used to attack Australian government services, six attacked Great’s Britain’s Royal Mail, one attacked Canada Post, and one attacked the Denmark Tax Authority.”

The NTIA recently published a proposal that would allow GoDaddy to redact registrant data from WHOIS registration records. The current charter for .US specifies that all .US registration records be public.

Interisle argues that without more stringent efforts to verify a United States nexus for new .US domain registrants, the NTIA’s proposal will make it even more difficult to identify phishers and verify registrants’ identities and nexus qualifications.

In a written statement, the NTIA said DNS abuse is a priority issue for the agency, and that NTIA supports “evidence-based policymaking.”

“We look forward to reviewing the report and will engage with our contractor for the .US domain on steps that we can take not only to address phishing, but the other forms of DNS abuse as well,” the statement reads.

Interisle sources its phishing data from several places, including the Anti-Phishing Working Group (APWG), OpenPhish, PhishTank, and Spamhaus. For more phishing facts, see Interisle’s 2023 Phishing Landscape report (PDF).’

Update, Sept. 5, 1:44 p.m. ET: Updated story with statement provided today by the NTIA.

Meet the Brains Behind the Malware-Friendly AI Chat Service ‘WormGPT’

WormGPT, a private new chatbot service advertised as a way to use Artificial Intelligence (AI) to write malicious software without all the pesky prohibitions on such activity enforced by the likes of ChatGPT and Google Bard, has started adding restrictions of its own on how the service can be used. Faced with customers trying to use WormGPT to create ransomware and phishing scams, the 23-year-old Portuguese programmer who created the project now says his service is slowly morphing into “a more controlled environment.”

The large language models (LLMs) made by ChatGPT parent OpenAI or Google or Microsoft all have various safety measures designed to prevent people from abusing them for nefarious purposes — such as creating malware or hate speech. In contrast, WormGPT has promoted itself as a new, uncensored LLM that was created specifically for cybercrime activities.

WormGPT was initially sold exclusively on HackForums, a sprawling, English-language community that has long featured a bustling marketplace for cybercrime tools and services. WormGPT licenses are sold for prices ranging from 500 to 5,000 Euro.

“Introducing my newest creation, ‘WormGPT,’ wrote “Last,” the handle chosen by the HackForums user who is selling the service. “This project aims to provide an alternative to ChatGPT, one that lets you do all sorts of illegal stuff and easily sell it online in the future. Everything blackhat related that you can think of can be done with WormGPT, allowing anyone access to malicious activity without ever leaving the comfort of their home.”

In July, an AI-based security firm called SlashNext analyzed WormGPT and asked it to create a “business email compromise” (BEC) phishing lure that could be used to trick employees into paying a fake invoice.

“The results were unsettling,” SlashNext’s Daniel Kelley wrote. “WormGPT produced an email that was not only remarkably persuasive but also strategically cunning, showcasing its potential for sophisticated phishing and BEC attacks.”

A review of Last’s posts on HackForums over the years shows this individual has extensive experience creating and using malicious software. In August 2022, Last posted a sales thread for “Arctic Stealer,” a data stealing trojan and keystroke logger that he sold there for many months.

“I’m very experienced with malwares,” Last wrote in a message to another HackForums user last year.

Last has also sold a modified version of the information stealer DCRat, as well as an obfuscation service marketed to malicious coders who sell their creations and wish to insulate them from being modified or copied by customers.

Shortly after joining the forum in early 2021, Last told several different Hackforums users his name was Rafael and that he was from Portugal. HackForums has a feature that allows anyone willing to take the time to dig through a user’s postings to learn when and if that user was previously tied to another account.

That account tracing feature reveals that while Last has used many pseudonyms over the years, he originally used the nickname “ruiunashackers.” The first search result in Google for that unique nickname brings up a TikTok account with the same moniker, and that TikTok account says it is associated with an Instagram account for a Rafael Morais from Porto, a coastal city in northwest Portugal.

AN OPEN BOOK

Reached via Instagram and Telegram, Morais said he was happy to chat about WormGPT.

“You can ask me anything,” Morais said. “I’m an open book.”

Morais said he recently graduated from a polytechnic institute in Portugal, where he earned a degree in information technology. He said only about 30 to 35 percent of the work on WormGPT was his, and that other coders are contributing to the project. So far, he says, roughly 200 customers have paid to use the service.

“I don’t do this for money,” Morais explained. “It was basically a project I thought [was] interesting at the beginning and now I’m maintaining it just to help [the] community. We have updated a lot since the release, our model is now 5 or 6 times better in terms of learning and answer accuracy.”

WormGPT isn’t the only rogue ChatGPT clone advertised as friendly to malware writers and cybercriminals. According to SlashNext, one unsettling trend on the cybercrime forums is evident in discussion threads offering “jailbreaks” for interfaces like ChatGPT.

“These ‘jailbreaks’ are specialised prompts that are becoming increasingly common,” Kelley wrote. “They refer to carefully crafted inputs designed to manipulate interfaces like ChatGPT into generating output that might involve disclosing sensitive information, producing inappropriate content, or even executing harmful code. The proliferation of such practices underscores the rising challenges in maintaining AI security in the face of determined cybercriminals.”

Morais said they have been using the GPT-J 6B model since the service was launched, although he declined to discuss the source of the LLMs that power WormGPT. But he said the data set that informs WormGPT is enormous.

“Anyone that tests wormgpt can see that it has no difference from any other uncensored AI or even chatgpt with jailbreaks,” Morais explained. “The game changer is that our dataset [library] is big.”

Morais said he began working on computers at age 13, and soon started exploring security vulnerabilities and the possibility of making a living by finding and reporting them to software vendors.

“My story began in 2013 with some greyhat activies, never anything blackhat tho, mostly bugbounty,” he said. “In 2015, my love for coding started, learning c# and more .net programming languages. In 2017 I’ve started using many hacking forums because I have had some problems home (in terms of money) so I had to help my parents with money… started selling a few products (not blackhat yet) and in 2019 I started turning blackhat. Until a few months ago I was still selling blackhat products but now with wormgpt I see a bright future and have decided to start my transition into whitehat again.”

WormGPT sells licenses via a dedicated channel on Telegram, and the channel recently lamented that media coverage of WormGPT so far has painted the service in an unfairly negative light.

“We are uncensored, not blackhat!” the WormGPT channel announced at the end of July. “From the beginning, the media has portrayed us as a malicious LLM (Language Model), when all we did was use the name ‘blackhatgpt’ for our Telegram channel as a meme. We encourage researchers to test our tool and provide feedback to determine if it is as bad as the media is portraying it to the world.”

It turns out, when you advertise an online service for doing bad things, people tend to show up with the intention of doing bad things with it. WormGPT’s front man Last seems to have acknowledged this at the service’s initial launch, which included the disclaimer, “We are not responsible if you use this tool for doing bad stuff.”

But lately, Morais said, WormGPT has been forced to add certain guardrails of its own.

“We have prohibited some subjects on WormGPT itself,” Morais said. “Anything related to murders, drug traffic, kidnapping, child porn, ransomwares, financial crime. We are working on blocking BEC too, at the moment it is still possible but most of the times it will be incomplete because we already added some limitations. Our plan is to have WormGPT marked as an uncensored AI, not blackhat. In the last weeks we have been blocking some subjects from being discussed on WormGPT.”

Still, Last has continued to state on HackForums — and more recently on the far more serious cybercrime forum Exploit — that WormGPT will quite happily create malware capable of infecting a computer and going “fully undetectable” (FUD) by virtually all of the major antivirus makers (AVs).

“You can easily buy WormGPT and ask it for a Rust malware script and it will 99% sure be FUD against most AVs,” Last told a forum denizen in late July.

Asked to list some of the legitimate or what he called “white hat” uses for WormGPT, Morais said his service offers reliable code, unlimited characters, and accurate, quick answers.

“We used WormGPT to fix some issues on our website related to possible sql problems and exploits,” he explained. “You can use WormGPT to create firewalls, manage iptables, analyze network, code blockers, math, anything.”

Morais said he wants WormGPT to become a positive influence on the security community, not a destructive one, and that he’s actively trying to steer the project in that direction. The original HackForums thread pimping WormGPT as a malware writer’s best friend has since been deleted, and the service is now advertised as “WormGPT – Best GPT Alternative Without Limits — Privacy Focused.”

“We have a few researchers using our wormgpt for whitehat stuff, that’s our main focus now, turning wormgpt into a good thing to [the] community,” he said.

It’s unclear yet whether Last’s customers share that view.