OpenAI backs binding UK AI rules as MPs call for a regulator

OpenAI backs binding UK AI rules as MPs call for a regulator

A UK parliamentary committee has called for an AI law and a single regulator with enforcement powers. The same day, OpenAI said it would back binding AI regulation for frontier developers, itself included. The business secretary warned against getting hyperbolic about the risks.

September 15, 2026 – 2:26 pm

Credit: The wub / CC BY-SA 4.0 via Wikimedia Commons

A cross-party committee of MPs and peers wants a dedicated UK AI law and a single regulator with enforcement powers. On the same day, OpenAI stated it would support binding UK rules for the few companies building the most powerful AI, which includes OpenAI.

Tom Duff Gordon, OpenAI’s head of policy for Europe, the Middle East and Africa, emphasized the need for “a stronger set of safety rules” in the UK. He suggested that the current political window should be taken advantage of by Prime Minister Andy Burnham’s government.

What the committee wants

The Joint Committee on Human Rights published its report, Human Rights and the Regulation of AI, on September 14th. It identified gaps in existing UK law regarding AI systems, particularly high-risk ones. The committee proposes:

• An AI bill with stricter obligations for higher-risk systems.
• A ban on specific uses considered incompatible with human rights.
• Mandatory transparency across the AI lifecycle.
• Establishment of an independent oversight body.
• Timetable for ratifying the Council of Europe’s AI convention.

The report highlights risks not covered by current law, including unsconsented facial recognition scanning and deepfakes used for sexual exploitation.

Alex Sobel, the Labour MP who chairs the committee, stated: "Nowhere in the world… has a current legislative and regulatory approach to AI that is fit for purpose."

The report also criticizes the limitations of the AI Security Institute (AISI), which tests frontier models but lacks regulatory authority.

What OpenAI is asking for

Duff Gordon outlined what OpenAI considers acceptable:

• Legislation focusing narrowly on national security and cybersecurity risks.
• Coverage only for a small number of labs and models at the frontier.
• Compulsory independent testing by third parties.
• Rules for monitoring and reporting safety incidents.
• AISI playing a central role.

As quoted separately in The Guardian, Duff Gordon stated: "We support stronger UK rules… for the handful of companies… developing the most powerful AI systems."