Google's latest Chrome update does more than patch ten security holes — it quietly marks a turning point in cybersecurity. In the October 15, 2026 release notes for Chrome 155, Google credited an Anthropic researcher "assisted by Claude" with discovering one of the fixed vulnerabilities, one of the first times an AI model has appeared in a browser's official security credits, as reported by Cyber Security News on October 17.
The update is available now for Windows, Mac, and Linux. Here's what was fixed — and why that AI credit matters.
What Chrome 155 Fixes
The stable channel update addresses ten vulnerabilities: one critical, six high-severity, and three medium-severity. Google is rolling out builds 155.0.8059.84/.85 for Windows and Mac and 155.0.8059.84 for Linux over the coming days and weeks. The good news: Google says none of these flaws are being exploited in the wild.
- CVE-2026-110922 (Critical): a use-after-free flaw in Chrome's Prerender feature, which loads pages in the background to speed up navigation. Reported by Xinyang Ge on October 2.
- CVE-2026-110913 (High): a memory corruption flaw in Chrome's DOM — the browser system that represents and manages web page structure. This is the one credited to "Xinyang Ge (Anthropic), assisted by Claude."
- Other high-severity bugs: five more flaws in V8, DOM, Core, and SVG, four of them also reported by Ge.
- Three medium-severity bugs: reported by Google's own researchers — improper input validation in Desktop, a use-after-free in Compositing, and memory corruption in Workers.
Why the "Assisted by Claude" Credit Is a Big Deal
Security researchers have used automation for years, but an AI model earning a name-check in Google's official credits is new. According to GBHackers' breakdown of Google's advisory, the DOM flaw was reported by Ge with Claude's assistance on September 29, 2026 — and it wasn't a one-off. The October 13 Chrome release carried the same "assisted by Claude" credit for CVE-2026-110169 in V8.
Two Chrome releases in a few days crediting AI assistance suggests that AI-assisted vulnerability discovery is moving from experiment to routine. Machines that can scan millions of lines of browser code without getting tired are finding bugs humans miss — and the window between a public fix and attackers understanding it keeps shrinking.
What Is a Use-After-Free Bug?
Two of the fixed flaws are use-after-free bugs, the same category as the critical entry. This happens when software keeps using a block of memory after it has been released. Depending on how an attacker can reach the flaw, it can cause a crash or allow memory corruption — and in some cases, arbitrary code execution inside the browser's renderer. These are exactly the kinds of bugs that make browsers attractive targets.
What You Should Do
The advice is simple:
- Update now: open Chrome's Settings → About Chrome to trigger the update check, then relaunch the browser.
- Don't wait for auto-update: the rollout happens over days and weeks, and the faster you patch, the smaller the window for attackers.
- Keep auto-updates on: Chrome usually patches itself, but a manual check after a security release is always worth it.
Google restricts access to detailed bug reports until most users have the fix, so the safest move is to be among the early updaters rather than the stragglers.
The Bigger Picture
The AI credit line is a tempo signal. As AI systems get better at both finding and exploiting vulnerabilities, the race between defenders and attackers accelerates. Google's October updates show the defensive side winning this round — ten flaws found, credited, and patched before anyone was attacked. But they also show that the era of AI-powered security research has officially begun, and browser users will feel its effects every time their software updates.
Disclaimer: Version numbers and severity ratings come from Google's published advisories and may be updated; check Chrome's release notes for the latest.



