Google's PageBreak agent finds 500+ XSS bugs with deterministic validation, near-zero false positives
rez0__ · x · 2026-09-30
Google's Product Security team details PageBreak, an internal agentic vulnerability discovery tool piloted in Nov 2025 and fully launched in Jan 2026 to autonomously test first-party web apps.
- Deterministic validation: instead of letting LLMs hypothesize bugs from code patterns, PageBreak verifies each candidate against a running environment, closing the loop on exploitability — yielding a near-zero false positive rate and avoiding the flood of "AI slop" reports.
- Scale and models: primarily powered by Gemini models (Gemini 3.1 Pro, Gemini 3.5 Flash), it has run at massive scale and uncovered over 500 XSS vulnerabilities.
- The project addresses the key bottleneck of LLM-based scanning: noise from unverified hypotheses that burdens product teams.
More from coding & agent
- Generic API-catalog MCP server vs one tool per job: a context-bloat experiment — MandeepKaur9407 · 2026-11-16
- Berkeley paper: LLMs answer with stale values even when the update is in context — ombre-ai · 2026-11-16
- skills-gate: one gate skill fronts 148 agent skills at just 389 tokens — Distinct-Pie2389 · 2026-11-16
- Open-weight Karpo-Decision-31B tops Decision Index at 73.9, one forward pass per answer — CodeByPoonam · 2026-11-16
- shadcn boosts @uisdk: one-shot UI rebuild that's instantly E2E-testable by AI — shadcn · 2026-11-16
- AI agent deletes 43,599 lines of OpenClaw test code in one PR — vincent_koc · 2026-11-16