White House quiet on OpenAI's Mythos-like model

OpenAI debuted GPT-5.5-Cyber, a cybersecurity model that rivals Anthropic's Mythos capabilities. The update achieved an 85.6% score in the internal CyberGym benchmark for reproducing software vulnerabilities. Despite its power, the release faced significantly less political pushback and White House scrutiny than Anthropic's Fable 5 and Mythos 5. This rollout may shift market dynamics by favoring broader access over regulatory-constrained competitors, raising questions regarding the Trump administration's specific concerns about rival AI models.