OpenAI Codex Security Scanned 1.2 Million Commits and Found 10,561 High-Severity Vulnerabilities Including OpenSSH and Chromium
The Hacker News / OpenAI·high signal
OpenAI launched Codex Security on March 21, an AI security agent that scanned 1.2 million commits across open-source repos in 30 days, finding 792 critical and 10,561 high-severity vulnerabilities in OpenSSH, GnuTLS, PHP, Chromium, and others. False positive rates dropped 50% across successive scans. Available in research preview to ChatGPT Pro/Enterprise/Business customers with free usage for the first month.