The EU's cybersecurity agency ENISA will meet Anthropic in San Francisco on Thursday, June 18, against the backdrop of tightened US export controls on advanced AI models. A European Commission spokesperson confirmed the planned meeting .
June 17, 2026 · EU–US AI Policy
US Export Curbs Cut Off Anthropic's Frontier Models — EU Demands Answers
After Washington ordered Anthropic to block foreign access to its most advanced models, the company disabled them worldwide. Now the EU's cybersecurity agency ENISA meets Anthropic in San Francisco to confront the fallout.
2
frontier models pulled: Claude Fable 5 & Claude Mythos 5
~3 days
from release (~Jun 9) to global cut-off
Global
disablement for all users, not just foreign nationals
How the directive escalated
US Commerce directive
National-security order to halt foreign access
→
Anthropic responds
Disables Fable 5 & Mythos 5 entirely, worldwide
→
EU impact
European users & institutions lose frontier access
→
Jun 19 meeting
ENISA & Anthropic meet in San Francisco
EU concern
Commission warns the curbs "should not be discriminatory." Loss of frontier access sharpens fears over European technological sovereignty.
Possible opening
Project Glasswing could give ENISA its first frontier-AI vulnerability-research access to Mythos — mirroring a parallel EU talk over OpenAI's GPT-5.5-Cyber.
The Thursday meeting — arranged at Anthropic's invitation and scheduled before the directive — becomes the venue to confirm practical responses after the regulatory shock, with no full Glasswing agreement yet reached.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…