May 2, 2026 GPT-5.5-Cyber: OpenAI Locks Away Its Cyber Model - After Mocking Anthropic for Doing the Same OpenAI GPT-5.5 Cybersecurity Anthropic Mythos AISI
