The EU wants to regulate AI but needs OpenAI and Anthropic to let regulators through the door

OpenAI has taken a significant step towards regulatory compliance by granting the EU Commission direct access to its new GPT-5.5 Cyber model for security review, signaling a proactive approach to AI oversight. In contrast, Anthropic has been less forthcoming, with multiple meetings yielding no access to its Mythos model, underscoring the challenges regulators face in establishing effective oversight in Europe.

The EU’s efforts to regulate AI are heavily reliant on the cooperation of major players like OpenAI and Anthropic. OpenAI’s willingness to collaborate may set a precedent, potentially easing the path for future regulations and fostering a more transparent relationship between tech companies and regulators. However, Anthropic’s reluctance raises concerns about the efficacy of voluntary cooperation, highlighting the need for a more structured regulatory framework that could compel compliance from all AI developers.

As the EU navigates this complex landscape, the outcomes of these discussions will be pivotal. If OpenAI’s model review proceeds smoothly, it could encourage other companies to follow suit, leading to a more standardized approach to AI regulation across Europe. Conversely, if Anthropic continues to withhold access, it may prompt regulators to consider more stringent measures to ensure compliance.

Looking ahead, the focus will be on how the EU adapts its regulatory strategies in response to the varying levels of cooperation from AI companies.