Agents
OpenAI Codex Security Research Preview: 1.2M Commits Scanned, 10,561 High-Severity Vulnerabilities Found Across OpenSSH, Chromium, PHP — 50% False Positive Reduction
OpenAI launched Codex Security in research preview — an AI security agent evolved from the Aardvark private beta — that has scanned over 1.2 million commits in the last 30 days, surfacing 792 critical and 10,561 high-severity findings across major open-source projects including OpenSSH, GnuTLS, Chromium, and PHP. The agent builds deep project context to surface complex multi-step vulnerabilities that rule-based scanners miss, while false positive rates have fallen more than 50% across scanned repositories over successive runs. Available free for one month to ChatGPT Pro, Enterprise, Business, and Edu customers via the Codex web interface.
↳ Follow the thread