Anthropic to offer EU access to its advanced Mythos model

Anthropic is now offering EU access to its advanced Mythos AI model, following months of cybersecurity concerns. The European Commission reported several productive meetings with Anthropic, and EU tech sovereignty spokesperson Thomas Regnier stated the bloc aims to better understand potential risks posed by powerful AI models. Mythos was initially launched in April to a limited group of companies under Project Glasswing, where it excelled at identifying software security flaws. The rollout triggered widespread concern about AI-driven cyber threats. Regnier emphasized that Mythos is part of a broader wave of emerging models, calling it a shared challenge requiring intensified collaboration with partners like the United States. Official terms of the EU access agreement remain undisclosed.