Google’s AI is digging up Chrome bugs that humans missed for years

Google is using Gemini-powered AI agents to scan Chrome’s codebase for security flaws, triage bugs, generate and review fixes, and write tests.

The system has already uncovered serious issues, including a sandbox escape bug that went unnoticed for more than 13 years.

Google says the approach saves significant developer time and speeds up Chrome security updates.

By shortening the time to verify and patch vulnerabilities, it reduces the window attackers have to exploit them.