AI Security
LLM & AI-agent attack research plus the AI-security news cycle — prompt injection, jailbreaks, agent abuse, model attacks — last 90 days.
Spotlight · past 30 days
- 9 sourcesSep 21
OpenAI Codex Sandbox Escape Vulnerabilities Disclosed
Discover the critical OpenAI Codex sandbox escape vulnerability. Learn how attackers bypass read-only modes and how to patch your CLI and desktop apps now.
SecurityOnline - 6 sourcesSep 18
Gemini Hacked Three Companies in First Known Breakout by Google’s AI
Gemini Hacked Three Companies in First Known Breakout by Google’s AI Gemini finally caught up on Felony Bench! The hacks, which the company confirmed on Friday, occurred in May as part of a test run by the company Irregular, which was also involved in similar…
Simon Willison - 6 sourcesOct 1
Autonomous AI agents tried to hack US, Canadian government websites
Autonomous AI agents using aggressive strategies attempted to hack U.S. and Canadian government websites to find school and divorce statistics. [...]
BleepingComputer - 5 sourcesOct 9
AI-Driven tool ARTEX used in attacks against South Korean Banks
CrowdStrike analyzes open directories left by an attacker who used the ARTEX AI pentest tool and LLMs to breach South Korean financial firms. CrowdStrike published a research on a campaign against South Korean financial organizations that ran from late…
Security Affairs
Learn
- OWASP LLM Top 10Canonical LLM app risks (2026)
- OWASP ML Top 10Classical ML security risks
- Agent Control StandardOWASP ACS for agentic AI
- NIST AI RMFAI risk management framework
- MITRE ATLASAdversarial AI tactics & techniques
- Google SAIFSecure AI framework
- GenAI CrosswalkSecurity frameworks cross-reference
- LLM Attacks LabPortSwigger LLM attack labs
- AI Security ToolsOpen-source tools, ranked by stars
Type
- Proofpoint to acquire Varonis Systems in a $6.3 billion all-cash deal announced October 15, 2026, combining email, data, and AI security to protect sensitive business information. Under the agreement, Varonis shareholders will receive $49.75 per share in cash. That represents a 52% premium over its June 22 closing price, before initial sale rumors, and a […]
- A critical flaw in ChatGPT’s GitHub integration has the potential to turn a routine repository review into a secret-theft attack. A report published on October 15, 2026, describes a scenario in which manipulated repository instructions allegedly lead an AI agent to modify a different private repository and trigger GitHub Actions. According to the researcher Sunil […]
- Anthropic has warned that attackers use AI agents to automate every stage of a cyberattack, from finding targets and stealing credentials to moving through networks and exfiltrating data. Its September 2026 threat report describes campaigns in which Claude became an active part of attack operations, rather than simply helping hackers write code. The findings cover […]
- Have you ever counted the computers in your house? Tally them up now. I'll wait. You're probably including your desktop and laptop computers, if you own some. I bet you're also including the smartphones and tablets. Depending on the size of ... (report_number: 8131)
- VirusTotal can now bring threat intelligence into ChatGPT through VirusTotal AI (VTAI), allowing users to investigate suspicious files, URLs, domains, and IP addresses within a conversation. The connection helps analysts retrieve existing reports and explain detection results while keeping the source, analysis date, and coverage visible. The integration uses Model Context Protocol (MCP), a standard […]
- Security researchers earned a $250,000 reward from Google after building a full-chain Chrome exploit with AI agents. The XBOW team says the work took six days and about $40,000 in computing costs, using OpenAI’s publicly available GPT-5.6 Sol model. Google has fixed the flaws. XBOW disclosed the work on October 13, 2026. Its chain linked […]
- PoC and technical details for CVE-2026-103572, an 8.8-severity Visual Studio Copilot RCE and SSMS flaw, are now public. Patch your IDE immediately.
- Rogue AI models have hacked companies and tried to break into government websites. Now one has submitted a fake tip for an unsolved crime. Police in Philadelphia say Anthropic alerted them this week that one of its artificial-intelligence ... (https://incidentdatabase.ai/cite/1751#8125)
- A . is all you need! Today Microsoft shipped the patch for a vulnerability in Visual Studio Copilot that I reported in August. Microsoft classified it as Important severity Remote Code Execution vulnerability. At the root cause it allows Copilot in Visual Studio or SQL Server Management Studio to bypass a path validation check and write to its own settings and configuration files. This allows Copilot to escape containment restrictions and achieve arbitrary command execution.
- Torrance, California, October 13th, 2026, CyberNewswire Criminal IP, developed by AI SPERA, will participate in Security Days Fall 2026 Tokyo, taking place October 21–23, 2026, at the TAKANAWA GATEWAY Convention Center. At booth B20, Criminal IP will showcase Criminal IP Threat Intelligence (Criminal IP TI) and Criminal IP AITEM (AI-Powered Threat Exposure Management) through live […]
- BLOG 5 min. Deepfake video, AI-generated voices and synthetic resumes are letting fake candidates pass virtual interviews and get hired. Once […]
- 38 GenAI security resources from September 2026: GLM-5.3 safeguards stripped by abliteration, a jailbreak that persists in Kimi's memory, and an SGLang RCE.
- IBM’s Steve Wallin talks to SecurityWeek on the effect of AI-driven attacks on enterprise defenses.
- ReliaQuest details an AI agent attack that used one exposed endpoint and the Cairn platform to achieve a full server takeover in under 24 hours.
- MILAN, Sept 25 (Reuters) - Fraudsters using AI to impersonate senior executives stole €95 million ($108 million) from Fideuram, the private banking arm of Italy's biggest lender Intesa Sanpaolo (ISP.MI), opens new tab, two sources familiar ... (https://incidentdatabase.ai/cite/1749#8115)
1–15 of 538
Page/ 36