The EU has begun enforcing the AI Act: first RFIs to model providers (tokenstead.ai)

The European Commission has taken its first formal enforcement step under the EU AI Act, launched requests for information (RFIs) to major general-purpose AI model providers, including OpenAI and Anthropic. This development, announced by Henna Virkkunen on August 29, 2026, comes just weeks after the Act's obligations became enforceable. The RFIs focus on model security, external evaluations, and market monitoring protocols, aiming to address growing concerns over the potential risks associated with AI models following a series of serious incidents earlier that summer. This enforcement action is significant for the AI/ML community as it marks the EU's commitment to regulatory oversight and accountability in AI deployment. Non-compliance with the RFIs could lead to heavy fines or even restrictions on models' availability in the EU market, highlighting the increasingly stringent regulatory environment. The RFIs also reflect the Commission's recognition of systemic risks tied to AI capabilities and underscore a stark contrast with the more lenient regulatory approach in the US, where the focus remains on voluntary cooperation. As the situation develops, model providers may face rising scrutiny, while self-hosted models currently evade direct oversight.