The UK does not have Artificial Intelligence-specific legislation or regulators, and instead regulates Artificial Intelligence through the context in which it is used. Existing laws already cover some applications, including the Online Safety Act 2023, while the government has also relied on non-statutory principles set out in its 2023 white paper on a pro-innovation approach. The UK has backed international safety efforts by signing declarations at the Bletchley and Seoul summits, and it created the Artificial Intelligence Safety Institute in 2023, later renamed the Artificial Intelligence Security Institute in 2025 to reflect a stronger focus on serious risks with security implications.
Labour came into office promising binding regulation for the handful of companies developing the most powerful Artificial Intelligence models, and that commitment was repeated in the 2024 King’s Speech. In March 2025, Feryal Clark, then minister for Artificial Intelligence and online safety, said the government was “continuing to refine its proposals and will launch a public consultation in due course”. But the government has not launched a consultation regarding cross-sector Artificial Intelligence regulation. Liz Kendall later indicated that the government was thinking in terms of targeted interventions rather than a single broad bill, citing measures in other legislation to address issues such as illegal chatbot content and children’s online experiences. The 2026 King’s Speech did not include a cross-sector Artificial Intelligence bill, although the Regulating for Growth Bill proposed an Artificial Intelligence Growth Lab to test products and regulatory reforms in live market environments, with successful reforms able to be embedded permanently into law.
The EU has taken a different route with the EU Artificial Intelligence Act, described by the European Parliament as the “the world’s first comprehensive AI law”. The act applies across sectors and uses a risk-based model. Systems with unacceptable risks are prohibited, high-risk systems face requirements such as risk management and human oversight, and lower-risk systems are subject to transparency duties. The act also covers general purpose Artificial Intelligence models, with obligations depending on whether they are used in high-risk settings, create systemic risks, or are released under a free and open licence. The act came into force in August 2024, with unacceptable-risk systems banned since February 2025, GPAI provisions applied since August 2025 and high-risk Artificial Intelligence compliance requirements were due to apply from August 2026. A recent ‘Digital omnibus on AI’ agreement has delayed the application of high-risk provisions to December 2027 or August 2028, depending on product type.
Stakeholder views remain divided. A 2024/25 survey by the Ada Lovelace Institute and the Alan Turing Institute found that 72% of people supported introducing Artificial Intelligence laws and regulations, up from 62% in 2022/23, while 89% supported an independent Artificial Intelligence regulator. The survey also found that 38% agreed and 48% did not agree that “it is important to keep up with other countries on AI, even if this means lighter rules”. The Ada Lovelace Institute argued that the UK needs a comprehensive legal framework, including powers to define safety, mandate pre-deployment testing for general purpose systems, and create stronger accountability and reporting obligations. Opponents of cross-sector legislation instead favor targeted regulation at the point of use, arguing that broad rules can create uncertainty, complexity, and barriers to innovation, a criticism also raised by business groups and companies responding to the EU framework.
