Google unveils PageBreak AI security agent that found 500+ XSS bugs
11/18/2026 — 11/18, 11:20·1 sources (all)·1 reports (all)
Key takeaways
- Google has introduced PageBreak, a Gemini-based security agent that has found more than 500 XSS vulnerabilities in its own web applications…
- unverified candidates stay internal to improve the verifier.
Background & analysis
On November 18, 2026, InfoQ reported that Google has introduced PageBreak, an AI-driven security agent built on Gemini. According to a paper published by Google, the agent has identified more than 500 XSS flaws across Google's own web applications since it began full operation in January of this year. Rather than guessing at defects from code patterns alone, PageBreak investigates potential vulnerabilities autonomously and then tests every hypothesis in a running environment using real exploit techniques. These checks include executing JavaScript to confirm XSS, writing to a globally readable location to confirm path traversal, or triggering outbound DNS requests to confirm remote code execution, or RCE.
The reporting process is strictly gated. PageBreak only submits a finding once a deterministic validator has successfully reproduced it against the live application, which Google says drives the false positive rate close to zero. Michał Bentkowski, a Google information security engineer, explained that the team deliberately prioritized deterministic verification from the early stages of development, closing the loop by validating candidate flaws in the running environment so that unverified reports never overwhelm product teams. Findings that fail validation are not shared with product teams at all; they stay internal, where they serve as leads for deeper future scanning, expose gaps in the agent's verification coverage, and guide the development of new validators. The reporting available so far does not describe any further public rollout or next steps for PageBreak.
AI-generated from 1 reports · updated 4 days ago
Related AI tools
- Heat index
- 25
- All sources
- 1
- All reports
- 1
- First seen
- 4 days ago
Reports on this story headlines open the original
Google says its Gemini-powered security agent PageBreak has found more than 500 XSS vulnerabilities across its web apps since going fully operational in January. A finding is reported only when a deterministic verifier reproduces the exploit in a running environment, which the company says keeps false positives near zero.
InfoQ 中文AI score 67
Other stories people are talking about
- Heat index 180SurgeSignal restricts mobile screen capture, citing AI agents as an existential threat4 sources
- Heat index 64Google Gemini Black Friday: AI Pro annual plan about $99.991 sources
- Heat index 33Google Rebrands Spark as Tasks, Shifts to Gemini 4 Hydrogen1 sources
- Heat index 223Gemini 4 Neon now live across Google platforms with expanded pricing4 sources
- Heat index 133Google adds GIF generation to Gemini, powered by Omni 1.2 Flash2 sources
How is heat calculated?About the methodHide
Heat counts how many independent sources covered a story in the last 48 hours: one source counts once no matter how many posts it published, decaying with a 24-hour half-life. What ranks first is what many people are talking about.
This page aggregates public feeds. Headlines and summaries are machine-organized and remain the property of the original authors; verify important facts at the source. If you believe a headline or summary infringes your rights, email the contact address in the footer with the page URL and basis of your claim, and we will remove or replace it after review.
- Surge
- Discussion rising fast
- New
- First report within 6 hours
- Rising
- Still gathering discussion

