OpenAI launches GPT-6 Cyber model for vulnerability discovery and fixes

[OpenAI] OpenAI introduced GPT-6 Cyber on November 5, 2026, describing it as its cybersecurity model for finding exploitable vulnerabilities and developing fixes. According to OpenAI, the model is trained to stay within authorized scope.