Google's PageBreak AI scanner confirms XSS bugs in running environments, finds 500+ with near-zero false positives
moyix · x · 2026-09-26
Google's Product Security team details PageBreak, an internal agentic web security scanner co-developed by researcher Michał Bentkowski (SecurityMB); XBOW researcher moyix notes the industry is converging on XBOW's philosophy of deterministic validation plus speculative AI findings.
Key points:
- Problem: LLM static analysis floods teams with "AI slop" false positives, making hallucination-vs-real-vuln triage the main bottleneck.
- Timeline: piloted November 2025, full project since January 2026; mostly powered by Gemini 3.1 Pro and Gemini 3.5 Flash.
- Design: agents verify candidate flaws against live running environments instead of guessing from code patterns, yielding near-zero false positives.
- Results: at scale it uncovered 500+ XSS vulnerabilities across Google's first-party web apps.
- Two companion blog posts share methodology and findings.
More from Safety
- Reprogramming fewer than 10 dormant neurons backdoors any LLM agent, no training needed — DimitrisPapail · 2026-11-11
- Running local open-source models? Dormant-neuron backdoors mean you may have let the right one in — DimitrisPapail · 2026-11-11
- Proving LLM backdoor-safety without revealing weights: ZKML over formal methods — DimitrisPapail · 2026-11-11
- LLMs can backdoor other LLMs by reprogramming a few dormant neurons to trigger rm -rf — DimitrisPapail · 2026-11-11
- OpenAI open-sources Codex Security CLI and SDK for AI-driven vulnerability scanning and fixing — openai · 2026-11-11
- Nando de Freitas: trusting models trained by others can be catastrophic, openness is essential for safe AI — DimitrisPapail · 2026-11-11