This past summer saw the first King’s Speech, which set out the government’s legislative agenda for the upcoming parliamentary session in the UK under the new Labour government led by Prime Minister Keir Starmer. The event marked an official indication that the new government would “seek to establish the appropriate legislation to place requirements on companies working to develop the most powerful AI models.” Beyond this, the new government has not announced any specific bills, but there is widespread speculation that it will follow up with legislation that resembles policies first laid out in the party’s manifesto.

The Labour government faces a public that is concerned with the risks associated with AI. The Centre for Data Ethics and Innovation (CDEI)’s August poll of over 4,000 members of the British public found that 45% of respondents viewed AI taking people’s jobs as one of the “biggest risks posed by the technology;” 34% believed loss in human creativity and problem-solving was one of the greatest risks. At the same time, it is also entering a rapidly evolving regulatory environment; many major jurisdictions have introduced both binding and voluntary regulation. This includes the European Union’s landmark EU AI Act and President Biden’s Executive Order on the Safe, Secure, and Trustworthy Development and Use of AI, which includes voluntary obligations for AI developers. Under former Prime Minister Rishi Sunak, the UK played a leading role in facilitating cross-jurisdiction, voluntary agreements between governments and AI companies, with a specific focus on AI safety and frontier models through its AI Safety Summit and the establishment of the world’s first AI Safety Institute. This approach also included a pro-innovation, laissez-faire position that excluded the introduction of binding regulation. Most observers expect Prime Minister Starmer to aim for a middle ground between the US and the EU, with a balanced approach that allows innovation to continue while creating binding requirements that address the technology’s risks. In this blog post, we examine what existing regulation and the regulatory pathways the new UK government can rely on to address AI policy.