GPT-6 Cyber: Advancing the Frontier of Aligned Cybersecurity Agents
OpenAI, Thursday, November 5th, 2026
OpenAI introduces GPT-6 Cyber, a security research model that finds and validates exploits while staying within authorized scope.
OpenAI announced GPT-6 Cyber, a cybersecurity model that scores 48.5% on the ExploitGym benchmark, up from 32.1% for GPT-5.6 Cyber, and that OpenAI classifies as Critical in cybersecurity under its Preparedness Framework.
Specialized alignment training kept it within authorized scope in 99.93% of honeypot tests without safety prompts and 100% with them.
OpenAI says the model found 81 new high-severity Firefox vulnerabilities and developed exploits for 948 Linux kernel vulnerabilities, with fixes disclosed to maintainers.
New safeguards include misuse monitoring, IP allowlisting and critical infrastructure protections evaluated with Dragos, and access starts with approved Daybreak Red customers through the API and Codex.