OpenAI Reportedly Used AI to Generate 948 Linux Privilege-Escalation Exploits
An automated pipeline developed by OpenAI reportedly generated 948 exploits targeting privilege escalation to root on Linux during a 30-hour run.
The reported run demonstrates how AI systems may be used to produce exploit code at scale. The available headline does not describe the pipeline’s architecture, the targets of the generated exploits, whether the exploits were independently verified, or whether any were responsibly disclosed or deployed.