After dissing Anthropic for limiting Mythos, OpenAI restricts access to Cyber

OpenAI has begun restricting access to its GPT-5.5 Cyber tool—a powerful cybersecurity utility capable of penetration testing and vulnerability exploitation—to credentialed security professionals only, despite CEO Sam Altman previously criticizing Anthropic for employing similar gatekeeping tactics with its Mythos tool. This dual approach reflects industry-wide concerns about AI-powered security tools falling into malicious hands, forcing IT leaders to navigate restricted access models and credential verification processes to leverage advanced security capabilities. The move signals that even as AI security tools mature, adoption will remain constrained by vendor policies and government coordination, requiring CIOs to plan accordingly for gradual rollout timelines and potential access barriers.