OpenAI's GPT-6 Cyber Hits Critical: 81 Firefox Flaws, 948 Linux Exploits, and a Tighter Leash
Posted on 6th Nov 2026 11:06:36 in Artificial Intelligence, Machine Learning
Tagged as: OpenAI, GPT-6 Cyber, Cybersecurity, Vulnerability Research, AI Safety
OpenAI has released GPT-6 Cyber, the first cyber model it rates Critical under its Preparedness Framework. It found 81 Firefox flaws and 948 Linux exploits, and stays in scope 99.93% of the time.
