After criticizing Anthropic for limiting Mythos, OpenAI also restricts access to Cyber

OpenAI will limit access to GPT-5.5 Cyber to verified critical cyber defenders, not the general public.

Users must apply and undergo credential review through OpenAI’s Trusted Access for Cyber program.

The model is intended for defensive uses such as penetration testing and vulnerability analysis.

OpenAI says it may broaden access later, with input from governments and other vetted users.