Pro-innovation approach to AI regulation

In short

The pro-innovation approach to AI regulation is the UK's policy, set out in a 2023 white paper, of governing AI through five shared principles applied by existing regulators, rather than through a single AI law or a new AI regulator.

Also called: AI regulation white paper, UK AI white paper, UK approach to AI regulation
Topic: Law, policy and the UK
Level: Some background helps
First described: Set out by the Department for Science, Innovation and Technology in the white paper presented to Parliament on 29 March 2023¹.

The white paper "A pro-innovation approach to AI regulation" was presented to Parliament on 29 March 2023¹. Instead of defining AI by a fixed list, it focused on two features that cause regulatory headaches: "adaptivity", which makes AI hard to explain, and "autonomy", which makes it hard to assign responsibility¹.

It set five cross-sector principles: safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress¹. Regulators such as the ICO, CMA, FCA and Ofcom apply them within their own areas. The principles were not put into law at first, and the paper said its central support functions "will not involve the creation of a new AI regulator"¹. It said a later statutory duty on regulators to "have due regard" to the principles might follow¹.

The government's response of 6 February 2024 kept this model. It gave £10 million to build regulators' AI skills, asked regulators to publish their AI plans by 30 April 2024, and accepted that targeted binding rules for developers of highly capable general-purpose AI would probably be needed in time².

This is the opposite design to the EU AI Act, which is one law with risk categories and its own enforcement structure³.

As of October 2026 the UK still has no general AI statute. The May 2026 King's Speech contained no specific AI bill⁴, and a September 2026 law firm note says a comprehensive AI bill is not expected until late 2027 at the earliest⁵. A private member's AI (Regulation) Bill from Lord Holmes of Richmond had a first reading in the Lords on 4 March 2025 but went no further⁶.

Example

A UK bank using AI to approve loans is not regulated by an AI law. Instead the FCA applies its existing rules and the ICO applies data protection law, each using the white paper's principles such as fairness and contestability¹.

Why it matters

It explains why the UK, unlike the EU, has no single AI law. If something goes wrong with an AI system here, the regulator for that sector is usually the one that deals with it.

Questions people ask

Does the UK have an AI law?

Not a general one, as of October 2026. AI is governed by existing laws and regulators using the 2023 white paper principles¹. No specific AI bill appeared in the May 2026 King's Speech⁴.

What are the five UK AI regulation principles?

Safety, security and robustness; appropriate transparency and explainability; fairness; accountability and governance; and contestability and redress¹.

How is the UK approach different from the EU AI Act?

The EU AI Act is one binding law with risk tiers, bans and duties for general-purpose models³. The UK instead asks existing regulators to apply five non-statutory principles within their own sectors¹.

Related terms

Sources

Back to the glossary