Google Discloses PageBreak AI Security Agent That Found 500+ XSS Bugs

| Source: InfoQ AI/ML

Tags: Google, PageBreak, Gemini, XSS, vulnerability detection, Michał Bentkowski

Google says its Gemini-powered PageBreak security agent has found more than 500 XSS bugs in its web apps since January, reporting only exploits that deterministic validators reproduce.

Details

In a published paper, Google describes PageBreak, an agent that uses Gemini models to investigate potential vulnerabilities in its web applications. Since entering full operation in January it has found more than 500 cross-site scripting bugs. Every hypothesis is tested against a running environment with a real exploit, such as executing JavaScript for XSS, writing to a world-readable location for path traversal, or triggering outbound DNS for RCE. Only validated findings go to product teams, giving a near-zero false-positive rate. Unverified candidates are kept internally to seed deeper scans and guide new validators. Google engineer Michał Bentkowski acknowledges that validator coverage is incomplete, which risks false negatives.