OpenAI announced Monday it's giving the European Union preview access to GPT-5.5-Cyber, a hardened variant of its latest model aimed at cyber authorities, governments and EU institutions. Anthropic, meanwhile, has held back on giving the EU access to Mythos, which dropped a month ago. The Commission has had four or five meetings with Anthropic but says discussions are at a 'different stage' than the OpenAI deal already on the table. With Anthropic also frozen out of the Pentagon's classified AI rollout earlier this month, the company is collecting an awkward pattern of regulatory access denials.
Feels like Anthropic is making a deliberate posture choice. Whether that posture survives the next funding round is another question.