Google’s Chrome Security team published a detailed account of how AI models have transformed their vulnerability management pipeline, and the headline figure is difficult to dismiss: in the last two Chrome releases alone, the team fixed 1,072 security bugs, more than the total fixed across the prior 23 milestones combined. That’s not a marginal improvement. That’s a different category of operation.
“In early 2026, we built an agent harness that used Gemini to find vulnerabilities across the broader Chrome codebase with higher efficiency and lower false positives. One of the bugs we found was a sandbox escape that would allow a compromised renderer to trick the browser into reading local files — a bug that quietly survived in our codebase for more than 13 years!” states Google. “For many of us, this moment cemented the potential of AI-powered vulnerability detection.”
The same tools that find bugs are now also triaging them. Historically, triaging a single security report took between 5 and 30 minutes of human time. The automated system now handles filtering duplicates, reproducing the bug, assigning severity, and routing it to the right team — the Chrome team estimates this is saving hundreds of hours of developer time per month.
Fixing bugs at scale required rethinking the entire repair pipeline. The current approach runs a fixing agent that produces multiple candidate patches, a separate critic agent that evaluates which is best, and a test-writing agent that verifies the fix works across all Chrome platforms before a human ever reviews it.
“We have partnered closely with Google DeepMind and Project Zero for years, including on BigSleep and CodeMender. These tools are natively integrated into our continuous integration (CI) system, running every 24 hours across all CLs to proactively detect security bugs.” continues the IT giant. “This integration has yielded significant results: in May alone, we blocked over 20 vulnerabilities from reaching production, including a critical S1+ issue.”
The external vulnerability reward program also saw a sharp increase: by March 2026, Chrome received more bug reports than it had in all of 2025, which prompted a change to the VRP to focus researchers on findings that add something beyond what internal AI tools are already catching.
The speed at which fixes reach users matters as much as the speed at which bugs are found.
“Once a fix has landed and is visible in the public open source codebase, attackers can start to reverse engineer and exploit the bug before the fix reaches users’ machines — so called “N-day” attacks.” states Google. “This is commonly referred to as the “patch gap.” Since fixes committed to the main “tree” typically take weeks to reach the Chrome Stable channel (what the vast majority of our users run), minimizing this patch gap is a critical part of our strategy.”
Even that cadence isn’t enough if users don’t restart their browser to apply the update, which most people delay for understandable reasons. Chrome is now working on dynamic patching that replaces background processes on the fly without requiring a full restart, and has already shipped a change in Chrome 150 that automatically restarts the browser on macOS when all windows are closed and an update is pending.
The longer-term structural work is about eliminating entire bug categories rather than patching individual instances. Chrome is expanding MiraclePtr, a technology that neutralizes use-after-free vulnerabilities, to more libraries including Skia, ANGLE, and Dawn, with a goal of covering up to 90% of UAF bugs on the GPU main thread. A separate “spanification” effort has already brought 97% of first-party Chrome code to compile cleanly with strict unsafe-buffer warnings — pushing spatial safety enforcement into the compiler rather than relying on runtime checks. The honest admission in the post is that runtime mitigations have a ceiling, and the long-term answer is rewriting high-risk components in Rust. Chrome is building a centralized Rust SDK to make that a routine engineering choice rather than a heroic one.
The scale of the dependency problem is worth stating plainly. Chrome has more than 2,300 third-party dependencies across Chromium and satellite projects, about 1,700 of which ship to users in products ranging from Android devices to cloud enterprise stacks. All of those are now being moved onto automated update pipelines that roll them to their latest upstream versions continuously, with safety signals from Google’s Open Source Security Intelligence Platform providing guardrails against introducing new problems in the process.
“The AI era has undeniably intensified the software security threat landscape, but by combining rapid deployment mechanisms with deep structural defenses, we are ensuring the advantage remains firmly with defenders. With this, Chrome and the broader web become safer with every update.” concludes the report.
Follow me on Twitter: @securityaffairs and Facebook and Mastodon
(SecurityAffairs – hacking, Google)