Urgent.News

What's breaking now, across thousands of outlets.

Tech

Security researchers spent 6 days doing their Chromework, built 1B user exploit

Xbow earned a $250K bug bounty and all the AI model got was a lousy t-shirt

Security researchers spent 6 days doing their Chromework, built 1B user exploit

Security researchers from startup Xbow utilized OpenAI's GPT-5.6 Sol model to exploit three zero-day vulnerabilities in Google Chrome, creating a full-chain exploit capable of executing code outside the browser's sandbox on Windows machines. The team believes this vulnerability affects around one billion users worldwide. Google compensated the Xbow team with a $250,000 bounty, their highest reward for a Chrome full-chain exploit.

The security researchers, comprising Brendan Dolan-Gavitt, Andres Luksenberg, Baul Lee, and Willem Pinckaers, spent just six days during Black Hat and DEF CON conferences, along with approximately $40,000 in token charges, to discover the vulnerabilities and construct the exploit chain. Xbow CEO and founder Oege de Moor expressed amazement at the efficiency of the team, stating that experts typically spend six months to achieve similar results.

The exploit chain, involving a V8 compiler bug, an animated-WebP decoder bug, and a memory-safety bug in Crashpad, initially allowed attackers to gain read and write access to V8's pointer-compression cage. The final stage exploited an overflow in Crashpad to cross the outer process boundary and achieve unsandboxed code execution on the host.

Google has since patched the vulnerabilities, but users must update to version 152.0.7977.82 or later to ensure protection. The researchers emphasized that the deployment of large language models (LLMs) like GPT-5.6-Sol has significantly lowered the bar for attackers, demanding enhanced cybersecurity measures.

Written by urgent.news from The Register's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Also reported by 1 other outlet

Read the original at theregister.com →

More in Tech

I replayed 7 repos' READMEs in fresh clones. 2 of them fail their own quickstart.

Your README is a promise nobody tests. Setup instructions rot silently. You rename a script, add an env var, switch a package manager, and the quickstart keeps promising a path that stopped working…

  • 2 out of 7 README files fail their own quickstart
  • npm run start command lacks corresponding script in package.json
  • coldstart-cli tool reveals hidden bugs in README documentation

Property Access Reviews: Readiness Health Endpoint Checks for Cached API Credentials

Use a cached credential-and-tier observation for readiness, and give that observation an explicit age limit. If the last confirmed observation is still fresh, return ready without contacting the…

  • Cached API credentials evaluated with observation age limit
  • Readiness health endpoint checks distinguish process availability from credential evidence
  • Degraded response indicates evidence beyond fresh window

What I Learned Mounting a WebDAV Share as My Default Working Drive in 2026

What I Learned Mounting a WebDAV Share as My Default Working Drive in 2026 For about three months I have been running with a WebDAV share mounted as my W: drive in Windows.

  • Author mounted WebDAV share as default working drive on Windows
  • Setup involved ScsDriver plugin ($5.5 lifetime) to controlled WebDAV endpoint
  • Drive remained available post-reboot without startup script

More from Wednesday 14 October →