❌

Reading view

This year’s bug count just passed 100K

Chalk up a new record for vulns. The number of Common Vulnerabilities and Exposures (CVEs) found and disclosed so far in 2026 surpassed 100,000 on Friday - more than double 2025’s total - and the reason can be summed up in two letters: AI. “It’s kind of wild,” Josh Bressers, VulnCheck’s director of coordinated vulnerability disclosure, told The Register. “The thing that almost scares the bejesus out of me is while 100K sounds like a lot, in a couple years we're going to say, remember when it was only 100K?” As of the time of our interview, about 4 pm UTC on Friday, VulnCheck counted 100,222 CVEs, he said. Microsoft alone added 1,094 security flaws to the total count on Tuesday, following 1,474 vulnerabilities the month before. “There's no denying the fact that vulnerability hunting with AI is driving a majority of what we're seeing.” All the major CVE Numbering Authorities (CNAs) - these are the organizations authorized by the CVE program to assign unique identifiers to security vulnerabilities - like VulnCheck, Microsoft, Oracle, GitHub, and other large enterprise IT vendors “are just screaming into the atmosphere right now,” Bressers said. “It’s AI.” Meanwhile, the percentage of these bugs that are exploited in the wild remains relatively consistent. About 1.4 percent (1,363) of the CVEs published this year have been added to VulnCheck’s Known Exploited Vulnerabilities (KEV) catalog. “But obviously, as the number of CVEs grows, [1 percent] of 1,000 is a lot different than [1 percent] of 100,000,” Bressers said. Plus, as other security researchers have documented, the window between the public release of a security patch and attackers developing working exploits – often by reverse engineering the patch while end users are still applying the fix – is rapidly shrinking. This is also thanks to AI. From CVE to KEV From January to June, 88 percent of the exploitation observed by CrowdStrike using public proof-of-concept (PoC) code occurred within 48 hours of the code's release. The security firm said China-linked groups such as Vault Panda and Genesis Panda moved even faster, launching attacks within 24 hours of disclosure. Similarly, VulnCheck found the median time from CVE publication to exploitation falling from 120 days in 2025 to 80 days during the first half of 2026. This window has closed even further during the second half of the year, which also seems to coincide with the AI-induced vulnpocalypse. “It’s often just hours,” Bressers said, noting 14 make-me-root vulnerabilities that VulnCheck disclosed on Friday in Copeland refrigeration controllers. Two are under active exploitation. These are the devices that Iranian hackers abused to spoil food and spark fires in supermarkets across Europe. “We published those CVEs, and literally minutes later, we were seeing hits on our honey pots,” Bressers said about the Copeland device vulnerabilities. “We’ve hit a point where with anything that looks interesting, someone is taking that information and seeing what they can do with it. Is it AI driven? Probably. It'd be hard to say it's not, because we've seen it really ramp up recently.” This influx of software bugs requires a sea change in how the entire industry approaches vulnerability disclosure and patching. “We don't have a security problem anymore,” Bressers said. “We have a data problem now. Instead of obsessing over vulnerabilities, we need to start looking at things like heat maps for infrastructure, and figure out the vulnerabilities that actually affect you. The CISA BOD helps with that.” What to prioritize in a patchpocalypse The US Cybersecurity and Infrastructure Security Agency (CISA) in June issued a cybersecurity directive that replaces flat-deadline vulnerability management policies. Instead of prioritizing patching based solely on CVSS severity scores, the directive instructs organizations to prioritize based on specific risk criteria, including whether the asset is internet-exposed, under active exploitation, and whether an attacker can automate the exploit. “It's telling people to figure out the vulnerabilities that actually affect you, and then put those at the top of the list, instead of just saying we're going to fix all the criticals,” Bressers said. AI-generated vulnerability disclosures also create huge headaches for open source software developers and maintainers, which Bressers describes as a security problem “super close to my heart.” “You've got individual developers that are drowning in vulnerability reports,” he said. “You have foundations being crushed by AI agents scraping their data or downloading their packages, millions of times a day, and it's straining the seams absolutely everywhere you look. The one thing I can say with certainty is: The way we currently do things has to change, and it's going to have to change fairly soon.” Different open source projects have taken different approaches to solving this issue. The Eclipse Foundation, which houses the Open VSX Registry, in April launched a “managed tier” for enterprise customers and other organizations running Open VSX at scale. Instead of using the free, open source version, the managed version guarantees a 99.95 percent uptime SLA backed by service credits. Its customers include AWS, IBM, Google, Cursor, and Windsurf. This paid option means commercial users who rely on the registry help fund its ongoing maintenance - and their own platforms’ security and performance. Meanwhile, curl’s maintainer, Daniel Stenberg, citing an “avalanche of vulnerability reports," updated its bug reporting rules last month. As of October 26, all low- and medium-severity reports go directly into the public GitHub repository, with only the high- and critical-flaws receiving private handling. Stenberg already shuttered curl's bug bounty program this year as AI-generated submissions piled work onto its security team. “A lot more needs to happen, and I think it is going to be a challenge to enact some of this change,” Bressers said. “When a package registry drastically changes something, people start freaking out, and the world is going to end, and people on LinkedIn are complaining - a lot of people are hesitant. No one wants to be the first to jump into the alligator-filled pool. That's one of our challenges.”®

  •  

Quantum Systems drones back in the air after malware scare

Quantum Systems says its drones are flying again with Germany's military after a cyberattack prompted a precautionary grounding. The company says it spotted the attack at the end of October and reported it to the relevant authorities, but disclosed few details about the intrusion. "The threat was contained thanks to the swift response of our IT and security teams in close and transparent collaboration with the authorities," it said on Thursday. "Through the quick measures taken and the characteristics of the incident, there is no risk in operating Quantum Systems' unmanned systems for the end users and the operators." It added that other players in the European defense industry, and those who support Ukraine's military efforts, have faced "persistent cyberattacks for years." The Bundeswehr grounded its Quantum Systems drones after malware was detected in an update to QBase, the company's QBase ground-control software, Der Spiegel reported, citing unnamed sources. QBase lets operators plan missions, monitor flights, and process data from the ground. The malicious update reportedly reached servers belonging to the German military and other customers, but did not directly infect the drones. "The fleet is operating normally around the globe – including Germany and the Bundeswehr," Quantum Systems told Reuters. "The Bundeswehr was the only customer to ground their systems as a precaution." A spokesperson for Germany's defense ministry declined to comment on the attack or the reported details. The Register asked Quantum Systems for further details. Founded in 2015, Quantum Systems has been providing drones to the Ukrainian war effort for years. Last month, Chancellor Friedrich Merz announced funding for "several tens of thousands" of the company's interceptor drones as part of Germany's winter aid package. The new drones will be built in Ukraine and support interceptor systems already in use on the front lines. Quantum Systems announced a $1.2 billion Series D funding round in July, valuing the company at around $8 billion. ®

  •  

Crackdown on UK medical record snooping after staff reported sneaking 300+ times this year

Britain's privacy watchdog has put healthcare bosses on notice after receiving more than 300 reports of staff snooping through medical records this year, demanding that executives personally confirm they're doing enough to keep prying eyes out of patient files. The Information Commissioner's Office (ICO) said on Thursday that it had written to more than 200 senior health leaders across the UK, warning that employees accessing records without good reason put patient privacy and public trust at risk. The ICO received and assessed more than 300 reports of medical record snooping between January and mid-October 2026. Staff were caught looking up people they knew, poking around patient files out of sheer curiosity, and sharing private information with people who had no business seeing it. The figure covers reports submitted by healthcare organizations rather than confirmed breaches, and excludes cases still awaiting assessment. ICO chief executive Paul Arnold wants health bosses to personally assure him that their organizations have safeguards in place to stop staff having a nose through records they aren't supposed to see. "Accessing someone's health records without a legitimate reason isn't harmless curiosity. It can cause enormous distress to patients and jeopardize the vital trust people place in our health service," Arnold said. In a three-page letter to chief executives, Arnold makes clear that simply trusting staff to behave themselves won't cut it. He wants to know who can access patient records, whether anyone is checking what they're looking at, and what happens when somebody gets caught snooping. Health bosses are also expected to investigate suspected breaches without dragging their feet, work out whether patients need to be told, and take action against staff caught accessing records they shouldn't. Executives have until November 27 to respond. Those who haven't reviewed their safeguards in the past six months, or aren't happy with what they find, have been given another three months to put their house in order. The problem isn't necessarily that healthcare workers are hacking into systems, as many already have legitimate access to patient records as part of their jobs. The trouble starts when they use that access to look up someone they're not treating, whether it's a neighbor, an ex-partner, or somebody whose name has appeared in the news. This isn't the first time health authorities have tried to stamp out the problem. In July, NHS England warned employees that accessing records without a legitimate reason could cost them their jobs or even land them in prison. That crackdown followed incidents involving staff accessing the medical records of victims of high-profile crimes, including the Nottingham attacks. NHS England also published guidance encouraging organizations to use access controls and audit systems to catch employees poking around where they shouldn't. Some newer electronic patient record systems can even flag suspicious access in real time, although the NHS acknowledged that monitoring capabilities vary depending on the technology organizations have installed. The scale of the problem has been troubling regulators for years. In July, the UK government told Parliament that 1,445 cases of inappropriate access had been reported to the ICO between 2019 and 2025. And the consequences can go beyond getting the sack. The ICO has prosecuted 16 healthcare workers for snooping through patient records over the past decade, with unauthorized access to personal information potentially amounting to a criminal offense under the Data Protection Act 2018. Arnold warned that the regulator could also take enforcement action against healthcare organizations whose protections aren't up to scratch. "I've written to chief executives across the UK asking them to personally assure me that their safeguards work in practice. Where they don't, we expect organizations to put that right," he said. With hundreds of reports already landing on the watchdog's desk this year, healthcare bosses now have a fortnight to explain how they're keeping staff out of records they have no business reading. ®

  •  

Brit police watchdog says cops' use of facial recog risks wrongful arrests

UK police forces must do more to prevent the wrongful arrests of people subjected to retrospective facial recognition (RFR) searches, according to a Criminal Justice Inspectorate that holds deep concerns about the way in which the technology is used. According to a report from His Majesty's Inspectorate of Constabulary and Fire & Rescue Services (HMICFRS), there is a large gap in understanding among police forces and the National Crime Agency about how RFR should be used in responsible policing. The inspectorate blamed "fragmented and underdeveloped" frameworks for governance and legislation. It pointed to a lack of bespoke legislation to provide a uniform approach to police facial recognition use – and said existing laws have not kept pace with the technology. HMICFRS also said the rules governing each force's use of RFR are mismatched, and called for a statutory code to be included in the forthcoming Police Reform Bill to set nationwide standards. Illustrating the breadth of misunderstanding across forces, the report noted that despite the message being routinely communicated by National Police Chiefs' Council (NPCC) reps, individual investigators "often" thought RFR results constituted evidence rather than intelligence. In many cases, they are unaware that RFR results require further enquiries before enforcement action can be taken. "We found that some investigators didn't always carry out enough corroborative enquiries before acting on potential matches," the report stated. "In a small number of cases, officers relied too heavily on RFR results, which led to inappropriate investigative action. "Although some forces had strengthened their processes after such incidents, practice remained inconsistent." Later in the report, the inspectorate described the findings as "deeply concerning." Some of the excuses police staff offered included acute demands on response teams leading to reduced opportunities to carry out further enquiries before acting on RFR matches. In one force, managers were known to encourage officers to "progress investigations quickly" after getting a match on the RFR system. However, the inspectorate clarified that just because an arrest is made, it does not necessarily mean it is wrongful, as investigators may need to rule out subjects from their inquiries. The onus for taking action should also be placed on the police forces themselves, given that some lack formal procedures regarding how RFR enquiries should be carried out, the report noted. The best forces had centralized facial recognition teams, which generally provided better governance and consistent processes. Officials fear long-term misuse of RFR, and other forms of facial recognition, will erode public trust in the technology, which it insists remains generally high. Surveys carried out in 2025 by the likes of the Ada Lovelace Institute and Alan Turing Institute, as well as the King's College London Policy Institute, showed a generally favorable view of facial recognition use in UK policing, with support ranging from 91 percent and 53 percent respectively. Home Office figures released on Thursday to complement the report show similar positivity toward the technology among police officers, with only 4 percent reporting a negative experience. The Register's comments section often paints a vastly different picture of public perception of facial recognition. Kathryn Stone, his majesty's inspector at HMICFRS, said facial recognition has transformed policing, allowing officers to compare images in seconds, allowing them to solve crimes more quickly, but echoed the report's main concerns. "Police forces must be able to show that appropriate safeguards are in place, risks relating to bias and data retention are effectively managed, and decisions are still subject to human judgment," she said. "A clearer statutory framework, stronger national standards and improved governance and oversight arrangements are essential to maintaining public confidence." As well as legislative and governance-related concerns, HMICFRS also noted in its report that officers' training around the use of RFR remains inconsistent, and if improvements were made, this would lead to more effective and lawful use of RFR across the UK. Another major issue was police forces' failure to meet their image retention obligations. Forces were found too often to retain custody images taken of suspects on the Police National Database, against which many RFR queries are made. Police guidance sets out the scenarios in which custody images can and cannot be retained, and the inspectorate found that forces have failed to delete images that should have been wiped. The Police Reform Bill includes a provision to group custody images with the likes of DNA and fingerprints, bringing about more robust measures to ensure they are retained or deleted in line with legal requirements. A Home Office review previously found in 2017 that introducing such a statutory framework was untenable. It concluded that police could not automate the deletion of these images, and doing so manually would be too time-consuming. Jasleen Chaggar, senior legal and policy officer at Big Brother Watch, said UK police cannot be allowed to continue to use RFR when it risks wrongfully arresting individuals. "The revelations that officers are treating facial recognition as bona fide evidence rather than a probability score is alarming," she said. "This approach significantly increases the risk that more innocent people will have officers knock on their door to arrest them for crimes committed in towns they've never visited, as we've seen before. "Worse still, police aren't even logging their use of retrospective facial recognition, meaning people could wrongfully become suspects in criminal investigations without any record of why they've been accused." Chaggar called on the government to "urgently step in" and help the UK's police forces to sort out the new frameworks needed for effective RFR use. ®

  •  

C++26 stumbles at finish line over Contracts dispute

The next version of the C++ programming language may be delayed at the last minute due to a dispute over a security-enhancing Contracts feature. Contracts have deeply split C++ technical maintainers. One group backs the C++ creator Bjarne Stroustrup, who is against the technology in its present form in grounds it was hastily implemented. But others are eager to see Contracts implemented, even arguing the tech's detractors are poisoning the C++ development process itself. The disagreement boiled over just before a vote to ratify the new version of the language. In October, the Draft International Standard (DIS) of C++26 failed its ISO ballot — the next-to-last stage in the International Organization for Standardization’s approval process during which a standard's technical details are formally ratified by a set of national standards bodies. The ballot saw 13 votes cast in favor, 9 against, and 12 abstentions, failing to meet the ISO requirement of having no more than 25 percent negative votes. Contracts was the major sticking point for those voting nay. Contracts dispute Contracts allow developers to add preconditions and postconditions to code, as well as assertions that check whether specified conditions hold. Such checks can help catch programming errors that might otherwise lead to undefined behavior. Design by Contract has long been a feature of languages like Ada and Eiffel. Developers have proposed adding Contracts to C++ for at least a decade, after when they were abruptly yanked from C++20 owing to concerns over dangerous defaults. But Stroustrup roasted the revised proposal (P2900) that made it into C++26. “The P2900 contracts and extensions aim to change the nature of C++ and are an existential threat to C++, adding complexity and bloat,” Stroustrup and his colleagues wrote in a paper last August. They noted that the implementation is messy, outdated, not sufficiently tested, and could be more effectively implemented as a library. Not everyone agrees. Andreas Fertig, a member of the C++ standardization committee, feels that the criticisms have already been vetted and that vetoing the proposal this late in the C++ release process is damaging to C++ itself. “The topic has been worked on for years and has received a majority vote multiple times in its current form. By voting ‘no,’ an established system, one based primarily on trust and respectful collaboration, has been shaken,” Fertig wrote in a blog post. The elephant in the room is, of course, the ongoing criticism of how easy it is for developers to accidentally create security holes in C++, leading more than a few projects to adopt Rust. Adding controls might, as Stroustrup notes, hamper C++’s chief attribute of being a blazing-fast language. Sore losers Fertig accused Stroustrup and Co. of being sore losers. “Every ISO standard is simply a set of compromises. Everyone loses sometimes, and everyone wins sometimes,” he wrote. But in this case, “the losing parties refused to accept their loss. The question that primarily determines how to move forward is whether the group can rebuild trust." Unlike Fertig, Harry Bott, CEO of the C++ Alliance – which maintains the open source EDG C++ compiler front end – does not see the Contracts discussion as disruptive to C++ itself. “The draft being sent back isn’t a failure of the process—it’s the process working as intended,” he wrote to The Register by email. “That doesn’t mean the process can’t improve. The goal should be to learn from this and strengthen the process so C++ can continue to evolve in a way developers can rely on.” Where to now? With the ballot failing at this late point of the standardization process – a rarity in C++ history – the language risks missing its expected shipping date of this year. That outcome may force everyone to refer to the new version as “C++27” rather than C++26. (That’s only the informal name; the official name of this eighth version of the language, should it be released next year, would be “ISO/IEC 14882:2027.”) ISO C++ working group WG21 is meeting next week in Búzios, Brazil, for the confronto final. According to a blog post from Thibault Ricord-Marchal, a software engineering consultant at the Paris-based Margo consultancy, the group faces a number of choices. They can carve Contracts out of the spec so the rest of C++26 can ship. That will be messy work, though. The undo button has to be smashed on a lot of library hardening already done using Contracts. Another option is moving Contracts into a separate technical specification, where work can continue to refine the technology. Or they could do a minor revision, vote again and hope the balmy Brazilian weather chills everyone out.®

  •  
❌