Anthropic grants EU cybersecurity agency access to Mythos 5 AI model

The European Union’s cybersecurity agency, ENISA, has obtained access to Anthropic’s Mythos 5, a highly advanced AI model specialized in cybersecurity. This access follows several months of negotiations between European institutions and the American AI lab.

Mythos 5 is noted for its sophisticated ability to identify and exploit computer vulnerabilities, which led to its initial availability being restricted primarily to select organizations in the United States. The European Commission confirmed that ENISA has begun testing the model to evaluate its capabilities and potential risks to information security.

This development follows previous access granted to ENISA for advanced models from OpenAI, such as GPT-6-ASTRA. Meanwhile, Anthropic continues to expand its public sector offerings, recently making the Fable 5.1 model available on the Claude for Government platform.