Chrome AI Vulnerability Discovered and Patched by Google

A new Chrome AI vulnerability has been patched after Google’s artificial intelligence system uncovered the flaw. The issue, tracked as CVE-2025-9478, involved a use-after-free bug in Chrome’s ANGLE graphics library. Without quick action, attackers could have executed arbitrary code across multiple platforms.

Discovery by Google’s AI

Big Sleep, Google’s AI-powered security agent, identified the vulnerability. Google developed the system with DeepMind and Project Zero to detect flaws before cybercriminals exploit them. Big Sleep previously identified a serious SQLite issue, proving that AI can strengthen defensive research.

Impact of the Vulnerability

The Chrome AI vulnerability carried high risk. By exploiting memory handling errors in GPU-accelerated content, attackers could trigger memory corruption and gain control. With Chrome serving billions of users across Windows, macOS, and Linux, the flaw presented a serious security concern.

Patch and Updates

Google has released fixes in Chrome version 139.0.7258.154 for Windows and macOS, and version 139.0.7258.155 for Linux. Users are urged to update immediately. The update closes the vulnerability and improves Chrome’s overall stability.

Why This Matters

The Chrome AI vulnerability highlights a turning point in cybersecurity. Artificial intelligence is no longer only a target—it is becoming a vital defense. Big Sleep shows how AI can uncover weaknesses at scale, potentially stopping attacks before they start. This shift could redefine how companies secure widely used software.

Security Recommendations

Users should take immediate action:

• Update Chrome to the latest version.
• Verify updates through Settings → Help → About Google Chrome.
• Enable endpoint protection tools to defend against future exploits.
• Stay alert as AI detection tools become central to cybersecurity.

Conclusion

The discovery and patching of the Chrome AI vulnerability prove the growing importance of AI in safeguarding technology. Google’s Big Sleep AI caught a flaw before attackers could weaponize it, showing the potential of machine intelligence in proactive defense. As threats evolve, AI is poised to play a central role in protecting billions of users worldwide.