OpenAI releases GPT-6 Cyber for cyber defense
OpenAI introduced GPT-6 Cyber, a model specialized for cybersecurity research.
- GPT-6 Cyber is intended for authorized security research within a defined scope.
- OpenAI reports gains in exploit testing and adherence to task boundaries.
- The launch begins with restricted access through Daybreak Red.
Context and analysis by Noosigma AI
Editorial summary and context based on the cited source. Claims and results attributed to the organization are reports from that source; they do not constitute independent validation by Noosigma AI.
The company is offering it to authorized professionals for finding vulnerabilities, reproducing exploits in controlled environments, and preparing fixes. Initial access will be limited to customers approved for Daybreak Red.
Why it matters
In OpenAI’s evaluation, the model scored 48.5% on ExploitGym, compared with 32.1% for GPT-5.6 Cyber. In a scope-following test, it avoided targeting out-of-scope services in 99.93% of cases without the deployment prompt and in 100% with it.
OpenAI also says the system helped identify 81 previously unknown high-severity Firefox vulnerabilities and produce exploits for 948 Linux kernel flaws. These figures are reported by the company, not an independent evaluation. Source: OpenAI, November 5, 2026.