Published
GPT-6 Cyber: OpenAI's AI Found 81 Firefox Flaws. Here's What to Update
OpenAI's new security model found dozens of Firefox flaws and built hundreds of Linux exploits in lab tests. Here is what that means and what you should update.
1. What it is
GPT-6 Cyber, launched November 5, 2026, is a restricted OpenAI model for finding and validating software vulnerabilities.
2. Firefox
OpenAI says it found 81 previously unknown high-severity Firefox flaws, many already patched by Mozilla.
3. Linux
Out of 1,724 reproduced kernel bugs it built working local privilege-escalation exploits for 948, in lab environments. These are not remote hacks.
4. Who can use it
Only approved Daybreak Red customers for now. OpenAI rates it Critical for cyber risk under its Preparedness Framework.
5. What you should do
Update Firefox from Menu > Help > About Firefox, and install your Linux kernel updates and reboot.
sudo apt update && sudo apt full-upgrade
Sources
OpenAI announcement; BleepingComputer; The Next Web, November 2026.
FAQ
Can I use GPT-6 Cyber?
Not yet. It is limited to approved security customers.
Is Firefox unsafe now?
No. Most of the flaws are already fixed. Keep Firefox updated.