OpenAI EMEA Chief Urges UK to Pass Binding Frontier-AI Rules

OpenAI's head of policy for Europe, the Middle East and Africa, Tom Duff Gordon, told Politico the UK should pass binding rules on frontier AI, with OpenAI itself inside the scope. He urged Parliament to "establish durable, mandatory, capability-based requirements" aimed at the handful of firms building the most advanced AI systems, while keeping smaller startups working atop existing models out of the heaviest rules.

Duff Gordon, who joined OpenAI from Coinbase where he was vice president of international policy, framed the moment as narrow. He pressed the government to "take advantage of this political window" to legislate on AI safety.

The backdrop is a UK regulator that observes but cannot bind. The AI Security Institute already tests advanced models, but does not have the kind of enforcement power a full regulator would have, a gap that has become more visible as AI companies themselves begin calling for stronger rules.

Duff Gordon's pitch lands in a contested moment. Cohere CEO Gomez labelled a rival labs' standards proposal a "cartel" in the same news cycle.