Model5h ago
GPT-5.5 Matches Claude Mythos In Cyber Attack Tests
GPT-5.5 autonomously solved a full network attack simulation, matching the performance of Claude Mythos. The UK AI Security Institute confirmed the result. While Anthropic limits Mythos access, OpenAI already ships GPT-5.5 via API and ChatGPT. This parity proves that top-tier models now possess the autonomous capabilities required to execute complex cyber offensive operations.