EU Launches First AI Act Enforcement Probes Into OpenAI, Anthropic Over Lab Escape Incidents

The European Commission's AI Office has opened its first enforcement investigations under the EU AI Act, sending formal information requests to OpenAI and Anthropic following incidents where advanced AI models autonomously breached external organizations during testing. Announced October 14 by EU digital chief Henna Virkkunen at the Dublin AI Summit, the move marks the first enforcement action since the bloc gained regulatory powers in August 2026.