--- title: Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps description: Google's internal AI security agent PageBreak has uncovered more than 500 cross-site scripting flaws across the company's own web applications by combining... canonical: https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps | daily.dev og:description: Google's internal AI security agent PageBreak has uncovered more than 500 cross-site scripting flaws across the company's own web applications by combining... og:url: https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti og:image: https://api.daily.dev/og/posts/goniJr5tI.png og:image:alt: Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps **[Dark Reading](https://daily.dev/sources/dr)** · 4 min read · 0 upvotes · 0 comments ## Summary Google's internal AI security agent PageBreak has uncovered more than 500 cross-site scripting flaws across the company's own web applications by combining LLM-driven vulnerability hypothesis generation with deterministic, non-AI validators that execute real exploit payloads to confirm findings before reporting them. The system, piloted last November and made a full product in January, is built primarily on Gemini 3.1 Pro and Gemini 3.5 Flash, though it can work with other models. Google plans to eventually connect PageBreak to CodeMender, its automated fix-generation system, so verified flaws can get proposed patches for engineers to review. Three example flaws, including a cache poisoning bug and an XSS issue, have already been fixed. Security experts quoted note the approach addresses a core industry problem: AI tools generating more potential vulnerabilities than teams can realistically triage, emphasizing that agents should not get the final say without independent validation. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is Google's PageBreak AI agent and how does it validate vulnerabilities it finds? PageBreak is an AI agent built by Google's Product Security team, running primarily on Gemini 3.1 Pro and Gemini 3.5 Flash, that scans first-party web applications for vulnerabilities. Instead of trusting the AI's hypotheses, it passes each potential flaw to a suite of specialized, non-AI-written validators that execute a real payload against the running application, confirming exploitability before reporting, achieving a near-zero false-positive rate. _Teams weighing AI-driven vulnerability scanning against false-positive overload can track approaches like this on daily.dev._ ### How many vulnerabilities has Google's PageBreak agent found so far and what kinds of flaws were they? Google's PageBreak agent discovered more than 500 cross-site scripting flaws across its own web applications. Specific disclosed examples include a cache poisoning bug in apis.google.com, an XSS flaw in admin.google.com, and insecure external handshakes in browser extensions, all of which have since been fixed. PageBreak began as a pilot in November and became a full product in January. _Developers evaluating agentic AppSec tooling can follow real-world results like these via daily.dev._ ### Will Google connect PageBreak to an automated vulnerability-fixing system? Yes, Google intends to eventually connect PageBreak to CodeMender, its automated vulnerability-fixing system, so that once PageBreak identifies and verifies a flaw, CodeMender can generate a proposed fix for product engineers to review and apply. This pairing aims to move beyond just finding bugs toward automatically proving, fixing, and verifying the fix worked. _Security engineers tracking the shift toward automated fix generation can follow updates through daily.dev._ ## Similar posts on daily.dev - [Cybersecurity Researchers Uncover Flaw in Google AI Coding Tool](https://daily.dev/posts/cybersecurity-researchers-uncover-flaw-in-google-ai-coding-tool-38fhs0zr4) · DevOps.com · 1 upvotes · 0 comments - [Best of 2025: Google Gemini AI Flaw Could Lead to Gmail Compromise, Phishing](https://daily.dev/posts/best-of-2025-google-gemini-ai-flaw-could-lead-to-gmail-compromise-phishing-2c7ar72su) · Security Boulevard · 0 upvotes · 0 comments - [Using AI agents to secure Google infrastructure](https://daily.dev/posts/using-ai-agents-to-secure-google-infrastructure-n2sosodh7) · Google Cloud · 0 upvotes · 0 comments --- Tags: [#security](https://daily.dev/tags/security), [#google](https://daily.dev/tags/google), [#google-gemini](https://daily.dev/tags/google-gemini), [#ai-security](https://daily.dev/tags/ai-security) [View this post on daily.dev](https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps","url":"https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti"},"datePublished":"2026-10-06T18:06:06.634Z","dateModified":"2026-10-06T18:08:16.612Z","description":"Google's internal AI security agent PageBreak has uncovered more than 500 cross-site scripting flaws across the company's own web applications by combining...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/dd8f225ee74ebb3512d6f28e0a8604e8?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/dd8f225ee74ebb3512d6f28e0a8604e8?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Dark Reading","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Dark Reading","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/dr","url":"https://daily.dev/sources/dr"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"security,google,google-gemini,ai-security","timeRequired":"PT4M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Dark Reading","item":"https://daily.dev/sources/dr"},{"@type":"ListItem","position":3,"name":"Google's PageBreak AI Agent Finds 500 Flaws in Its Web Apps"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/google-s-pagebreak-ai-agent-finds-500-flaws-in-its-web-apps-gonijr5ti#faq","mainEntity":[{"@type":"Question","name":"What is Google's PageBreak AI agent and how does it validate vulnerabilities it finds?","acceptedAnswer":{"@type":"Answer","text":"PageBreak is an AI agent built by Google's Product Security team, running primarily on Gemini 3.1 Pro and Gemini 3.5 Flash, that scans first-party web applications for vulnerabilities. Instead of trusting the AI's hypotheses, it passes each potential flaw to a suite of specialized, non-AI-written validators that execute a real payload against the running application, confirming exploitability before reporting, achieving a near-zero false-positive rate. Teams weighing AI-driven vulnerability scanning against false-positive overload can track approaches like this on daily.dev."}},{"@type":"Question","name":"How many vulnerabilities has Google's PageBreak agent found so far and what kinds of flaws were they?","acceptedAnswer":{"@type":"Answer","text":"Google's PageBreak agent discovered more than 500 cross-site scripting flaws across its own web applications. Specific disclosed examples include a cache poisoning bug in apis.google.com, an XSS flaw in admin.google.com, and insecure external handshakes in browser extensions, all of which have since been fixed. PageBreak began as a pilot in November and became a full product in January. Developers evaluating agentic AppSec tooling can follow real-world results like these via daily.dev."}},{"@type":"Question","name":"Will Google connect PageBreak to an automated vulnerability-fixing system?","acceptedAnswer":{"@type":"Answer","text":"Yes, Google intends to eventually connect PageBreak to CodeMender, its automated vulnerability-fixing system, so that once PageBreak identifies and verifies a flaw, CodeMender can generate a proposed fix for product engineers to review and apply. This pairing aims to move beyond just finding bugs toward automatically proving, fixing, and verifying the fix worked. Security engineers tracking the shift toward automated fix generation can follow updates through daily.dev."}}]} ```