Google's Product Security team describes PageBreak, an internal agent mostly using Gemini models that hunts vulnerabilities in Google's first-party web applications and only reports findings confirmed by non-AI validators against running applications. Google reports over 500 XSS vulnerabilities found with near-zero false positives, while apps on its high-assurance web frameworks yielded only 2 XSS bugs as of 4 September 2026.
It shows a concrete design for suppressing AI-generated false positives. Google also reports that apps built on its secure-by-design frameworks yielded very few bugs to the agent, though that comparison is an uncontrolled self-report.
Key facts
As stated in the sources, with where to find them.
- PageBreak began as a pilot in November 2025 and became a full project in January 2026; most usage relies on Gemini 3.1 Pro or Gemini 3.5 Flash.Section 'PageBreak'
- It uncovered over 500 XSS vulnerabilities across Google first-party web applications.Section 'PageBreak'
- As of September 4, 2026, only 2 XSS vulnerabilities were found across hundreds of apps built on Google's high-assurance web frameworks.Section 'PageBreak vs a High-Assurance Framework'
- Unverified candidate findings are kept as seeds for later scans and are not sent to product teams.Section 'Internal Feedback Loop'
Findings that cite this record
No tracked finding cites this record yet.
Key questions this bears on
- Is AI shifting the balance between finding and fixing vulnerabilities?Discovery is ahead. AI finds real vulnerabilities faster than they are fixed, and simple checks overstate how often AI patches work.