OpenAI says the gated model scored 48.5% on ExploitGym, helped build 948 Linux exploits and found 81 new high-severity Firefox flaws.
By [Ryan Merket](https://runtimewire.com/author/ryan-merket)
· Published
· Updated
Primary source: [x.com](https://x.com/OpenAI/status/2118390190246879554)
Why it matters #
OpenAI is putting a model it classifies as Critical for cyber capability behind an approved-customer gate. The test for defenders, and for the startups selling around frontier models, is whether gated exploit development produces patches faster than the same capability spreads outside controlled programs.
OpenAI introduced GPT-6 Cyber on November 5th, a cybersecurity model the company says is more capable at finding and fixing vulnerabilities and more reliable at staying within boundaries set by researchers. OpenAI (@OpenAI) announced the model in a thread on X, and OpenAI described it in its GPT-6 Cyber announcement as "a major leap in security research".
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.