A security team used a public AI model to assemble a chain of Chrome bugs into a full-system exploit in six days, pushing browser defenses into the spotlight.
A security team used a public AI model to chain three Chrome bugs into a full-chain exploit in six days, potentially letting attackers run code on Windows machines just by loading a web page.
Google fixed the bugs and XBOW received compensation under the vulnerability rewards program. About 1 billion Windows Chrome users were at risk before the fixes, and the AI sprint cost roughly $40,000 in compute.
What this means for you is simple: keep Chrome updated and restart the browser after updates to stay protected.
The clock was short and the stakes are high: researchers used publicly accessible AI to stitch together three Chrome bugs into a full-chain exploit in six days, potentially letting an attacker run code on Windows machines just by loading a page. The flaws bypass Chrome’s layered sandboxing, allowing code to run outside the browser after a page load. Google has confirmed the bugs and issued fixes across the Chrome update line tied to that release window.
Beyond speed, the episode shows how AI can accelerate vulnerability research. The XBOW team says the AI-driven sprint cost about $40,000 in compute and happened during the Black Hat and DEF CON conferences in Las Vegas. The researchers had been working on the problem for months, framing the project as a challenge they set in January and building toward a rapid six-day sprint later on.
For everyday users, the most important takeaway is simple: keep Chrome and your operating system up to date. If you don’t restart Chrome after an update, you may miss critical protections in the latest fixes. And remember, the exploit could have affected any Windows user who had the page loaded before the patches landed.
The attack linked three separate bugs, each in a different Chrome component: memory handling inside the JavaScript engine, the WebP image decoding path, and the crash handler that runs outside the sandbox. All three bugs were triggered by simply loading a webpage, and Google patched them in the corresponding update cycle.
Google’s vulnerability rewards program pays as much as a quarter of a million dollars for a full-chain exploit. XBOW notes this was a notable use of the program and a sign that AI-assisted research is reshaping how fast attackers could assemble powerful chains. The program caps these awards to a small number per year.
In plain terms, this story highlights two practical steps: update Chrome promptly when you see a new version, and restart the browser after the update finishes to ensure the latest protections are active. The attack did not require user interaction beyond visiting a page, so staying current is a key defense for casual browsing too.
Security teams and browser makers will continue refining defenses against AI-assisted research. Expect focus on faster patch cycles, stronger sandboxing, and possibly new reward structures that deter weaponizing AI tools. For users, that translates to staying vigilant about updates and watching for official safety advisories after major fixes.
A team used a public AI model to chain three Chrome bugs into a full-chain exploit that could run code on Windows via a loaded web page. Google fixed the flaws and XBOW received compensation under the rewards program.
About $40,000 in AI compute for the six-day project.
You're reading the quick version.
Read next
Amazon ramps to 27 satellites a week as Leo nears launch
Amazon says its Leo constellation is weeks from offering commercial broadband after building its 1,000th satellite and accelerating production.
Hirudo malware secretly swaps legitimate banking apps with patched copies that bypass in-app security checks, letting thieves steal credentials and spy on devices.
A congressional inquiry into dating platforms finds scammers prospering for years, with calls for faster defenses and a central scam-reporting portal.
U.S. prosecutors say North Korean operatives recruited Americans to help breach U.S. companies, with authorities seizing hundreds of laptops in Texas as part of the investigation.
The Daily Brief
Every morning, the news that matters from AI, phones, apps and the people shaping tech. Explained in plain English. Free.
One email a day. No spam, unsubscribe anytime. Privacy policy
