When AI Builds the Exploit: OpenAI’s LPE Pipeline and the Patch Velocity Imperative
OpenAI’s 30-hour agentic pipeline turned 2,308 Linux kernel vulnerability candidates into 948 working LPE exploits at $1...
Read More81 articles
OpenAI’s 30-hour agentic pipeline turned 2,308 Linux kernel vulnerability candidates into 948 working LPE exploits at $1...
Read MoreNorth Korea’s TraderTraitor stole $6.3 billion from Binance — but the dollar figure undersells the story. The October 20...
Read MoreAnthropic is activating SynthID-Text watermarking on its 4-series API models today, November 3, 2026, with the EU AI Act...
Read MoreOn October 29, a malicious build of the official Claude Code VS Code extension delivered a persistent SSH backdoor to de...
Read MoreThree posts, three organizations, three different failure modes — agent trust boundaries, third-party software exposure,...
Read MoreAmazon’s new SDE role guideline says outcomes matter more than AI tool adoption. The context behind it — a 13-hour outag...
Read MoreJohns Hopkins Bayview notified 3,182 patients after a worker uploaded a PHI spreadsheet to an unapproved public AI tool....
Read MoreCathedral, an AI-powered cybersecurity startup, gained access to Anthropic’s internal meeting transcripts, salary data, ...
Read MoreAI agents have already reached government health data portals, probed major health system perimeters using attacker trad...
Read More