OpenAI says its new GPT-6 Cyber has already uncovered thousands of exploits

OpenAI has introduced GPT-6 Cyber, a cybersecurity AI model that has already identified thousands of exploits in critical software like Linux and Firefox, according to the company's official announcement. The new model reportedly delivers a 51% improvement over GPT-5.6 Cyber on the ExploitGym cybersecurity benchmark. OpenAI's GPT-6 Cyber is exclusively accessible through its Daybreak program, a cybersecurity initiative only available to verified individuals and organizations.

Initially, the AI is limited to a select group of Daybreak Red customers, with access gradually expanding to all Daybreak Red users. OpenAI aims to expand Daybreak to include advanced security research for various organizations, independent researchers, and open-source maintainers.