NVIDIA and 36 other organizations have formed the Open Secure AI Alliance to develop and share open technologies, techniques, and tools for securing software and artificial intelligence (AI) agents. The 37-member group spans cloud, security, enterprise software, and AI companies, including Microsoft, Cisco, Cloudflare, CrowdStrike, Hugging Face, IBM, Palo Alto Networks, Red Hat, and the Linux

Dysphoria, an Internet of Things (IoT) botnet line tracked by CNCERT and XLab, has adopted blockchain-based name services and infected-device relays after a March law-enforcement operation against JackSkid infrastructure. The researchers say the design makes the botnet harder to disrupt. CNCERT, China's national computer emergency response team, and XLab, the threat-intelligence lab of Chinese

Public exploit details released on July 27 show how an unauthenticated request can reach PHP's eval() function inside vBulletin and execute code on an unpatched forum server. The attack requires no account, administrative access, or interaction from another user. SSD Secure Disclosure lists vBulletin 6.2.1 and earlier, and 6.1.6 and earlier, as affected, but does not give a lower version

Monday starts with the usual promise that everything is under control. Then the logs wake up. This week, trusted tools crossed lines, old flaws found new work, exposed systems stayed exposed, and attackers kept hiding inside normal-looking services. Nothing looked strange at first. That helped. That is the mood. Here is the full recap. ⚡ Threat of the Week OpenAI Says Its AI Agent Went Rogue

n8n has patched a high-severity expression-sandbox escape that could let an authenticated workflow editor execute operating-system commands on the server running the automation platform. Security Joes found the flaw while probing n8n's February fix for CVE-2026-27577 for another bypass. The affected ranges are <2.31.5 and >=2.32.0,<2.32.1. n8n fixed the flaw in versions 2.31.5 and

Cybersecurity researchers have flagged a Microsoft Teams-themed phishing campaign that employs "secure document" lures to deliver legitimate remote monitoring and management (RMM) tools. "The victim was directed through compromised web infrastructure to a counterfeit Microsoft Store page claiming that Microsoft Teams had to be updated before the shared document could be opened," ZeroBEC said in

The China-linked cybercrime group behind the use of income tax-related phishing lures targeting Indian taxpayers, tax professionals, and corporate finance teams has been observed using a sophisticated crypter service called Cruciferra. According to a new analysis by Proofpoint, Cruciferra has been utilized by various unrelated cybercriminal threat clusters to deliver a wide array of remote

Cybersecurity researchers have flagged fresh malicious cyber activity by a threat actor with ties to East Asia targeting government entities in the Middle East. The intrusions have resulted in the deployment of previously unreported malware families dubbed TELESHIM, MIXEDKEY, and BINDCLOAK, according to Zscaler ThreatLabz. The cybersecurity firm said it detected the campaign earlier this month.

GitHub has announced a new cooldown mechanism in Dependabot, allowing the tool to wait at least three days after a release is published before opening a pull request. "The cooldown configuration option in the dependabot.yml still controls the behavior, though, so you can choose a different cooldown parameter that fits your project," the Microsoft-owned subsidiary said. According to GitHub, the

A malvertising operation dubbed SourTrade is making victims' browsers build the final Windows executable themselves, using a legitimate Bun runtime as its base instead of serving one complete malicious file from a fixed URL. Confiant, which detailed the campaign on July 23, 2026, said it has operated since late 2024 and impersonated TradingView, Solana, and Luno to target retail traders and

Security firms ThreatBook and Imperva say attackers are targeting a critical flaw in Fastjson, Alibaba's JSON library for Java. In affected Spring Boot applications, a malicious JSON request can execute code without authentication, with the privileges of the Java process. Tracked as CVE-2026-16723, the vulnerability carries an Alibaba-assigned CVSS score of 9.0. The confirmed chain requires

Security researchers at depthfirst published working exploit code on July 24 for a GitLab flaw that GitLab patched six weeks earlier, on June 10. It runs commands as git on any self-managed 18.11.3 server that has not taken the update. Any authenticated user who can push to a project can run it. The attacker commits a crafted Jupyter notebook and opens its commit diff, which leaks a heap

For years, phishing campaigns targeting financial institutions followed the same playbook. Victims were tricked into entering usernames and passwords, attackers collected the credentials, and accounts were compromised later when an opportunity arose. That model is changing. Recent investigations into insurance-focused phishing operations reveal a more immediate approach. Instead of harvesting

Threat actors linked to the Cl0p (aka Chubby Scorpius, FIN11, Graceful Spider, and Lace Tempest) ransomware campaign are exploiting flaws in internet-exposed PTC Windmill and FlexPLM deployments as part of a new data extortion campaign. "Attackers chain a pre-authentication information disclosure in the FlexPLM WSDL endpoint with a server-side flaw in the Windchill login servlet, enabling

The operators of the DevMan ransomware-as-a-service (RaaS) scheme are maintaining a dedicated web platform that offers affiliates the ability to build payloads, oversee earnings, and manage various aspects related to victims. Swiss cybersecurity company PRODAFT is tracking the centrally administered RaaS operation under the name Funky Mantis. "The portal combined build generation, finance,

The North Korean threat actors behind the ClickFix-style campaigns that employ typosquatted Zoom and Microsoft Teams domains have been found to operate an active phishing kit to impersonate the videoconferencing platforms in social engineering campaigns designed to deliver malware. "BlueNoroff has operationalised trust abuse by combining compromised industry contacts, social engineering, wallet

Researchers H0j3n and Aniq Fakhrul published a working exploit on July 24 that lets a low-privileged Active Directory user obtain a certificate for a Domain Controller and authenticate as that machine. They codenamed the flaw Certighost. Because Domain Controller accounts carry directory replication rights, the resulting Kerberos credential can retrieve the krbtgt secret through DCSync.

Cybersecurity researchers have disclosed a critical vulnerability in OpenAI's ChatGPT Workspace Agents that could have allowed a single phishing link to stealthily build, authorize, and deploy an autonomous artificial intelligence (AI) agent inside a victim's organization. The vulnerability has been codenamed AgentForger by Zenity Labs. The issue has since been addressed by OpenAI as of June 8,

A crafted SVG submitted to Bing's image search ran commands as NT AUTHORITY\SYSTEM on Microsoft's production image-processing workers, and as root on the Linux machines in the same fleet. XBOW's testing got the same result on workers across different hosts and network ranges, so the problem sat in Bing's image tier, not on one bad machine. Microsoft issued two critical CVEs, CVE-2026-32194 and

AI agent security is moving through a familiar maturity curve: adoption, then visibility, and finally, control. But what we've collectively discovered is that enforcing least privilege for AI agents is harder than we ever imagined. This is why there are so many approaches, from prompt filtering to identity-layer access controls. Where we've collectively landed is that understanding the intent of