
Google's Chrome Security team utilized AI models to identify and patch 1,072 security vulnerabilities across the last two browser releases, exceeding the total fixes from the previous 23 milestones.
The automated system, which includes Gemini-powered agents, now handles bug triaging and patch testing, saving hundreds of hours of developer time per month.
In May 2026 alone, the integration of these AI tools blocked over 20 vulnerabilities from reaching production, including a critical S1+ issue.