OpenAI Identifies 948 Linux Kernel Privilege Escalation Vulnerabilities Using Autonomous AI Model
OpenAI has announced the discovery of 948 local privilege escalation vulnerabilities within the Linux kernel. The findings were produced by an upcoming autonomous cyber security model that scanned the latest version of the kernel as of mid-October 2026. In addition to identifying these flaws, OpenAI's system autonomously generated corresponding patches for the bugs. To ensure accuracy and safety, the company stated that all AI-generated patches underwent review by human researchers before being finalized.
Topics: AI-driven security Automated patching Linux kernel vulnerabilities OpenAI