OpenAI researchers used a network model to identify and patch 948 local privilege escalation vulnerabilities in the Linux kernel in a single run.
OpenAI security researchers built an autonomous pipeline for vulnerability discovery, exploitation, and remediation. Author Tibo shared the findings, noting the team will share more work over the coming months. In a single run, an upcoming network model generated patches for 948 local privilege escalation vulnerabilities in the Linux kernel, with total API costs of $1.2 million. A funnel chart shows the process started with 2,308 candidate discoveries and yielded 291 functional PoCs.
Return to original text
A short introduction is retained here; the full text from the source website is not reposted. For specific wording and subsequent updates, please refer to the original article.
Open AIHOT original