Google’s AI Hacker Finds 500+ XSS Flaws and Builds Working Exploit Chains
Google's PageBreak agent verified over 500 XSS flaws, including exploit chains, in its own web applications.
Google disclosed PageBreak, an internal AI security agent that verified more than 500 cross-site scripting flaws in its own web applications. Piloted in November 2025 and run as a full project since January 2026, it uses Gemini models to propose issues and a browser-based validator to confirm that injected JavaScript actually executes. Notable chains included cache poisoning of a JavaScript file server that could have affected apis.google.com, an admin-console redirect completed via a signature endpoint, and a universal XSS in the Tag Assistant extension. Google reported no evidence of attacker use and is pairing confirmed findings with automated patching.
- PageBreak uses Gemini to propose flaws and a browser validator to confirm execution.
- Google verified more than 500 XSS issues; the pilot began in November 2025.
- A JavaScript cache-poisoning chain could have affected apis.google.com; no attacker use was found.
- Separate chains hit the admin console and the Tag Assistant extension.
- High-assurance frameworks showed only two XSS bugs, both limited endpoints.
- Vendors
- Products
- PageBreakTag Assistant
- Organizations
- AI models
- Gemini
From the original article
4 min read at cybersecuritynews.com
Google has revealed an internal AI security agent that found more than 500 verified cross-site scripting, or XSS, flaws across its own web applications. The system, called PageBreak, searches for weaknesses that could let an attacker run unwanted code in a visitor’s browser, including on sensitive services. The announcement matters because AI scanners can generate […]
By Tushar Subhra Dutta · cybersecuritynews.com
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.