OpenAI Codex Scans 1.2M Commits, Flags Over 10K Critical Vulnerabilities

OpenAI’s new Codex Security assistant processed more than 1.2 million code commits across diverse projects, automatically surfacing 10,561 high‑severity flaws. By ingesting full project context, the AI model identifies complex bugs that static analysis often misses, and it generates concrete remediation patches while aiming to keep false‑positive rates low for developers.

For defenders, the results demonstrate that AI‑driven code review can dramatically accelerate vulnerability discovery in the software supply chain. Integrating such tools into CI/CD pipelines gives security teams early visibility into risky changes, improves prioritization of remediation effort, and reduces the window of exposure before attackers can exploit newly introduced defects.

Categories: AI Security & Threats, Vulnerabilities & Exploits, SOC & Automation

Source: Read original article