Research
GPT‐6 Cyber: Advancing the frontier of aligned cybersecurity agents
First reported openai.com
Page published
Earliest dated coverage: 5 Nov 2026 · First observed: 5 Nov 2026 · Latest dated coverage: 5 Nov 2026
Coverage timeline
Single-source research — one report is available.
Why it matters
GPT‑6 Cyber’s reported exploit-development capabilities and scope-compliance results underscore the importance of independently testing authorization boundaries before deploying autonomous security agents.
OpenAI’s GPT‑6 Cyber announcement reports improved automated exploit development, with a 48.5% ExploitGym score versus 32.1% for GPT‑5.6 Cyber. A separate alignment evaluation reportedly found that the model avoided exploiting out-of-scope services in 99.93% of tests without its deployment prompt and 100% with the prompt.