Does the UK have an AI Act equivalent?

What is the UK approach in outline?

The UK approach, as set out in the 2023 White Paper on a pro innovation approach to AI regulation and in the government response that followed, is principles based and regulator led. Existing regulators apply five cross cutting principles, safety, security and robustness, appropriate transparency and explainability, fairness, accountability and governance, and contestability and redress, within the areas they already oversee. There is no central AI regulator and no UK equivalent of the EU AI Office, although the main digital regulators coordinate through the Digital Regulation Cooperation Forum. The approach is deliberately different from the EU horizontal statute model. It leaves AI governance largely with domain regulators that already understand their sectors, supported by central coordination functions. That difference has practical consequences. A UK company cannot look for a single UK AI statute that mirrors the AI Act chapters, and cannot look for a single UK AI authority that mirrors the AI Office. It must identify the regulators that oversee its sector, the law that governs its activity, and the guidance that each regulator has published for AI within its remit.

What happened to the anticipated AI Bill?

An AI Bill for the most advanced models, sometimes described as a frontier models bill, was anticipated in reporting before the 2024 general election and in the period that followed, but, as checked on 29 October 2026, no AI Bill was before Parliament and no standalone AI statute had been enacted. The government Blueprint for AI regulation published on 21 October 2025 effectively replaced the anticipated AI Bill as the near term legislative vehicle in the reporting checked for this guide. The Blueprint centrepiece is the AI Growth Lab, a national programme of issue specific regulatory sandboxes in which particular rules can be temporarily relaxed for licensed pilots, intended to let promising AI uses be tested under supervision before wider rules are finalised. The Blueprint also included a public call for views on whether the programme should be run in house or overseen by regulators. Reporting checked for this guide also described a Regulating for Growth Bill announced in a later King Speech as a vehicle for sandbox powers, with timing uncertain and the Bill not yet introduced. Those later developments must be checked on official parliamentary and government sources, because secondary reporting on unintroduced legislation is not a reliable basis for a legal conclusion. This guide does not assert the current parliamentary position beyond the checking instruction.

For the broader regulatory framework, see non-EU and international overview.

What is the AI Growth Lab proposal?

The AI Growth Lab, as announced on 21 October 2025 by the Secretary of State for Science, Innovation and Technology, is a programme of sector specific sandboxes where businesses can test AI products in real conditions, with some rules temporarily relaxed under licence and with safeguards in place. The initial sectors described in the announcement included healthcare, transport and robotics in advanced manufacturing, with examples including AI tools to support health workers and to speed planning approvals. Participants would operate under time limited licences, with monitoring by regulatory and technical experts, and with suspension of testing where risks are unacceptable. The Lab is a proposal and a programme, not a statute, and its legal basis, its licensing powers and its relationship to existing regulator sandboxes must be checked on GOV.UK and on the regulator sites. A UK company should not treat participation in a sandbox proposal as an authorisation, and should not treat the Lab as equivalent to an EU AI regulatory sandbox under Articles 57 to 59, which has a statutory basis, a sandbox plan and an exit report. The two sandbox concepts share a name and a testing idea, but they operate under different legal frameworks.

What law does apply to AI in the UK?

AI in the UK is governed by existing law, including data protection law, equality law, consumer law, product safety law, financial services law, health and medicines law, online safety law and sector specific law, each applied by the regulator that oversees that area. Data protection is the most generally applicable framework, through UK GDPR and the Data Protection Act 2018, as amended by the Data (Use and Access) Act 2025, with the Information Commissioner as regulator. Automated decision making, profiling, and the information, representation, human intervention and contest rights that attach to significant automated decisions must be checked in the current UK GDPR provisions and ICO guidance, because the 2025 Act changed the automated decision provisions and the ICO consulted on draft guidance in 2026. Sector law adds further duties: financial services through the Financial Conduct Authority, medical devices through the Medicines and Healthcare products Regulatory Agency, online safety through Ofcom, competition through the Competition and Markets Authority, and equality through the Equality and Human Rights Commission. A UK company must therefore map its activity to the laws and regulators that govern it, rather than searching for a single AI law.

How should a UK company handle EU AI Act duties?

A UK company that places an AI system or a general purpose model on the Union market, or whose output is used in the Union, is in scope for the AI Act under Article 2(1) and must meet the AI Act duties for those activities, in addition to UK law for its UK activities. The company should keep two records: a UK record that maps its UK law, regulators and guidance, and an EU record that maps its AI Act scope, classification, role, authorised representative and conformity position. The records should be kept distinct, even where the same system and the same evidence are used for both, because the legal bases, the authorities and the duties differ. The company should appoint an authorised representative established in the Union where Articles 22 or 54 require it, and should check whether its UK establishment, its Union customers or its Union output use trigger the duty. It should not assume that UK compliance, or UK regulator guidance, satisfies AI Act duties, and it should not assume that AI Act compliance satisfies UK duties. Each framework must be met on its own terms, with local advice where the position is borderline.

Continue with UK sector regulators and AI.

What does this guide not decide?

This guide does not decide the current UK parliamentary position, does not reproduce UK statutes or regulator guidance, and does not give legal advice for any UK activity. The UK position described here is the position as checked on 29 October 2026 from secondary reporting summarised in the sources note, and it must be rechecked on GOV.UK, on parliamentary sources and on regulator sites before it is relied on. UK AI policy, legislation and guidance develop, and a guide that states a UK position without a checking instruction would quickly become out of date. A company with a specific UK activity should obtain UK legal advice on the current law and guidance for that activity, and should confirm the AI Act position separately with a competent authority or legal adviser.

Frequently asked questions

Continue through this AI Act cluster

• non-EU and international overview
• UK sector regulators and AI
• territorial reach in Article 2
• third country providers and authorised representatives
• output used in the Union test
• providers