OpenAI Urges UK Lawmakers To Regulate Frontier AI

OpenAI is now asking the UK to regulate the most powerful AI labs, including itself, while keeping smaller startups out of the heaviest rules.

The company told British lawmakers that the government should use the current political window to introduce legislation for the handful of firms building the most advanced AI systems. Tom Duff Gordon, OpenAI’s head of policy in Europe, said the UK should act now, according to The Guardian. OpenAI’s position is that regulation should focus on frontier labs, not smaller companies working with less powerful models.

That is a careful argument. OpenAI wants oversight, but not broad rules that could slow the whole AI startup market. It is asking lawmakers to distinguish between a company training frontier models and a startup building a customer service tool on top of existing models. That distinction may make sense, but it also gives the biggest labs a role in shaping the rules that govern their own category.

The call comes as a UK parliamentary human rights committee warned that current law is not ready for the potential consequences of advanced AI. The committee pointed to risks around public face-scanning, explicit deepfakes, workplace monitoring and decisions that could affect people’s rights without proper human oversight.

The UK’s AI Security Institute already tests advanced models, but it does not have the kind of enforcement power that a full regulator would have. That gap is becoming more visible as AI companies themselves begin calling for stronger rules. Anthropic has asked for a coordinated slowdown in frontier development, while OpenAI, Google DeepMind and Elon Musk have signalled support for parts of that idea.

This puts the UK in a tricky position. Move too slowly and it risks becoming a place where powerful models are deployed before oversight catches up. Move too aggressively and it may push investment and talent elsewhere. OpenAI is trying to frame the middle ground as targeted regulation for only the largest model developers.

TechBooky has already looked at why AI rivals now agree the frontier may need a speed limit. The UK story shows what comes next. Once companies admit that safety rules are needed, governments will ask who writes those rules, who enforces them and what happens when a lab decides the next model is too important to delay.