OpenAI has introduced GPT-5.4-Cyber, a variant of its latest flagship model tuned for defensive cybersecurity work, and is limiting access to vetted users through an expanded Trusted Access for Cyber program. The release puts a more capable model in the hands of security teams while keeping distribution tightly controlled.
OpenAI on April 14 expanded its Trusted Access for Cyber program and introduced GPT-5.4-Cyber for vetted defenders, a move that broadens access to more permissive security tooling while keeping the rollout limited to verified users and teams.
Anthropic is limiting access to its new Mythos Preview model to a small group of technology and cybersecurity companies, a sign that frontier AI firms are tightening distribution as they court security buyers and confront misuse concerns.
