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

OpenAI announced a limited rollout of GPT-5.5 'Cyber' to vetted cyber defenders via an application process; the toolkit is described as supporting penetration testing, vulnerability discovery/exploitation, and malware reverse engineering, prompting concerns about potential misuse and drawing comparisons to Anthropic's similarly restricted Mythos which reportedly experienced unauthorized access.