OpenAI Restricts Cyber Access After Mocking Anthropic

Sam Altman previously criticized Anthropic for limiting access to its cybersecurity tool Mythos, calling it fear-based marketing. Now OpenAI is doing the same with its competing tool, Cyber, built on GPT-5.5. OpenAI will roll out Cyber to vetted cybersecurity professionals through an application process, citing concerns about misuse. The tool can perform penetration testing, vulnerability identification, and malware reverse engineering. OpenAI says it is working with the U.S. government to expand access to legitimate users.