OpenAI to Preview GPT-6 Cyber Security Model in Coming Days

The cybersecurity framing matters more than the model version number: positioning GPT-6 Cyber as a security tool, paired with a deployment product, suggests OpenAI is courting enterprise security teams rather than general consumers. That puts it on a collision course with established security vendors and, more subtly, with the offensive-security community that will test its guardrails immediately. The deployment layer is the real tell — automation plus safety controls is how vendors sell into regulated buyers. Whether the security-specific model proves materially different from a general frontier model, or mainly a packaging and policy exercise, is the open question.