King's Speech promises regulation of 'the most powerful AI technologies'

But no specific AI bill

The King's Speech on Wednesday mentioned measures to enhance the legal safeguards around the most powerful AI technologies, but stopped short of announcing an AI bill.

Labour's manifesto outlined plans to "ensure the safe development and use of AI models by introducing binding regulation on the handful of companies developing the most powerful AI models and by banning the creation of sexually explicit deepfakes", while also saying its "industrial strategy supports the development of the Artificial Intelligence (AI) sector."

These efforts include removing restrictions to the building of new datacentres, and creating a National Data Library to centralise public data and research programmes in order to better deliver data-driven public services.

In his speech to parliament, the King said the government "will seek to establish the appropriate legislation to place requirements on those working to develop the most powerful artificial intelligence models."

Although meat is still to be put on the bones, and there was no specific mention of an AI bill (instead there is to be a digital information and smart data bill) the UK's focus on the largest LLMs represents a reset of the strategy of the previous government, which favoured a light-touch self-regulatory approach.

Most legal experts expect Labour's strategy to sit somewhere between the Sunak government's laisse faire approach, and the more prescriptive EU AI Act. The latter, which comes into force at the beginning of next month, divides AI systems into four risk categories, banning the most dangerous outright, and placing stringent requirements on creators and distributors of high risk systems, which include autonomous vehicles, robots and industrial automation, HR systems and those used for medical diagnoses.

The EU AI Act also includes a set of rules including transparency requirements for the largest general purpose AI (GPAI) models (meaning foundation models like GPT4), which may be in line with the UK's plans.

Reaction to the King's Speech

Amanda Brock, CEO of OpenUK welcomed the focus on the largest AI models: "Our Labour government is well-advised to tread carefully with its AI bill capturing those who empower the most powerful AI, ensuring not to close market opportunities or limit UK innovation. The bill's focus being appropriately narrow and from a safety lens, regulating only the largest Models, is well received," she said.

"For the UK to stay relevant globally, and to build successful AI companies, openness is crucial. This will allow the UK ecosystem to grow its status as a world leader in open source AI, behind only the US and China."

However Silkie Carlo, CEO of Big Brother Watch, said on X: "AI regulation is welcome but 'the most powerful AI models' are a limited part of the picture. We'll campaign for an AI Bill to protect rights from everyday hidden AI already affecting the public - in welfare, public services, facial recognition, policing"

The fact that AI regulation was mentioned in the Kings Speech means it's likely to be fast-tracked, said Jonathan Armstrong, partner at Punter Southall Law.

"Labour plans to set up a new Regulatory Innovation Office which will look at the challenges of AI and support existing regulators including the Information Commissioner's Office and the Competition and Markets Authority in using their existing powers to regulate AI. ... With a decent majority and a favourable legislative timetable it may even be possible for new UK legislation to come into force by the same time as the EU legislation in 2026."

But Dan Reavill, head of technology and commercial transactions at law firm Travers Smith, bemoaned the lack of mention of a specific AI bill: "This was an opportunity for the new government to make its mark in the world of AI regulation and to play 'catch-up' with the EU. It remains to be seen whether the UK will continue to follow the Tory government's approach, and take a largely hands-off approach to AI regulation, or whether we might see some actual AI laws coming down the pipeline."

Anita Schjøll Abildgaard, CEO and co-founder of startup Iris.ai, also urged the government to align its legislation with the EU: "AI brought up so early in the Labour government clearly shows AI regulation as a top priority. This is a positive step, recognising the urgency and importance of governing this rapidly evolving technology. The UK, however, need not reinvent the wheel. The EU AI Act has already made significant strides in balancing innovation with responsible oversight. By aligning its approach with these efforts, the UK can not only promote global coherence in AI governance but also enhance international collaboration."