AI BriefingOpenAI Discovers 948 Linux Kernel Privilege Escalation Bugs Using AI
·2026.11.05 09:00
Key point
OpenAI's autonomous vulnerability discovery prototype identified 948 valid local privilege escalation exploits in the Linux kernel at a cost of $1.2 million in API credits.
Details
- OpenAI's AI prototype found 948 valid local privilege escalation vulnerabilities in the Linux kernel, allowing unprivileged users to gain root access.
- The process involved identifying 2,308 candidate vulnerabilities, dynamically reproducing 1,724, and generating candidate patches for maintainer review.
- The discovery effort consumed $1.2 million USD in API credits.
- OpenAI's GPT-5.5 Cyber model found more than 5x more novel local privilege escalations than the GPT-5.6 Sol model, despite three months of kernel hardening between scans.
- Researchers predict frontier models will likely discover existing vulnerabilities in software projects within the next few years, emphasizing the need for high-assurance software and secure code training.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
OpenAI launches Images 2.6 with 50% faster generation and GIF supportCat Wu joins OpenAI to lead product management for Core Products