Anthropic initially released Project Glasswing to about 50 partners in April to test the model for cybersecurity flaws.

Anthropic is negotiating EU access to its Mythos AI cybersecurity model, which can find zero-day vulnerabilities and generate exploits autonomously.

ENISA becomes the first EU agency to join Anthropic's Project Glasswing, gaining access to the Mythos AI model that found 10,000+ zero-day vulnerabilities in critical software.