OpenAI Restricts GPT-6 Cyber to Vetted Security Customers
The model found 81 previously unknown Firefox flaws, while OpenAI assigned it the highest cybersecurity rating in its safety framework.
Key takeaways
- OpenAI says GPT-6 Cyber identified 81 previously unknown, high-severity Firefox vulnerabilities.
- The model helped develop working exploits for 948 Linux kernel vulnerabilities, including 588 previously unknown flaws.
- Access will expand to all approved Daybreak Red customers over the coming weeks.
OpenAI has released GPT-6 Cyber, a model for vulnerability research and exploit development, to a small group of approved customers. The company says it found 81 previously unknown Firefox vulnerabilities that Mozilla classified as high severity.
OpenAI said the model also helped researchers develop working exploits for 948 Linux kernel vulnerabilities, including 588 previously unknown flaws. The company said it disclosed 304 critical Linux findings to maintainers, along with proposed patches.
OpenAI gave GPT-6 Cyber a “Critical” cybersecurity rating under its safety framework, up from “High” for GPT-5.6 Cyber. The company said it is seeing more attempts to misuse AI, including by malicious cyber actors and groups seeking to distill its models.
The new model is available to select customers of Daybreak Red, OpenAI’s access tier for advanced authorized security work. Access is expected to expand to all approved Daybreak Red customers over the coming weeks; GPT-5.6 Cyber became available to all of them on Thursday.
This article was produced with the help of AI technology. Based on reporting by the outlets listed above.
