OpenAI Model Patched 948 Linux Kernel Flaws Autonomously

OpenAI’s unreleased cyber model autonomously found, validated, and patched 948 Linux kernel privilege escalation flaws for $1.2 million.