OpenAI showcases an autonomous vulnerability discovery and remediation pipeline, generating 948 patches for local privilege escalation in the Linux kernel in a single run.

OpenAI security researchers have built an autonomous pipeline for discovering, exploiting, and patching network vulnerabilities. Using a newly launched cybersecurity model, they verified and generated patches for 948 local privilege escalation (LPE) vulnerabilities in the Linux kernel in a single run, with total API costs reaching $1.2 million. Funnel data shows that out of 2,308 candidate findings, 1,724 passed verification, resulting in 948 usable LPE exploits and 291 usable PoCs.