OpenAI says an upcoming cyber model's pipeline produced 948 Linux kernel privilege-escalation exploits

OpenAI describes an autonomous pipeline that finds, reproduces, exploits and patches vulnerabilities in the Linux kernel, run with an upcoming cyber model that has not been released. In a mid-October run against a stock desktop Linux distribution, OpenAI says the pipeline analyzed 2,308 candidate findings, dynamically reproduced 1,724 vulnerabilities, and produced 291 information-leak proofs of concept and 948 working local privilege-escalation (LPE) exploits, which let an unprivileged local user gain root. Human researchers reviewed the results before reporting them to maintainers, together with candidate patches.

OpenAI puts the run's cost at $1,213,430 in API credits, or $1,279 per LPE exploit, and calls that an upper bound. On an adapted kernelCTF environment with reduced attack surface, it says the pipeline found 21 LPE exploits for about $32,936. It also says the agents used at least 10 novel Linux exploitation techniques, and that the new model found more than five times as many novel LPEs as a GPT-5.6 Sol run in July.

OpenAI says over 100 of the Linux LPEs were patched or had staged patches from other contributors within 10 days, with about 300 more being discussed publicly. It advises that a low-privileged account on an unpatched system be treated as equivalent to root access against a well-resourced attacker, and recommends sandboxing such as Firecracker and gVisor. The figures and conclusions are OpenAI's own.