AI regulation

Horizon scanning

UK AI regulation – more delays to the UK's AI Bill?

What is happening?
On 29 April 2025 the Minister for Science, Lord Vallance, indicated during a parliamentary debate on the government's plans to introduce cross-sector legislation on artificial intelligence (AI), that the government will launch a consultation "later this year".

This is an important update from the government because, as Viscount Camrose put it during the debate, "the uncertainty for everybody affected by AI, whether in the tech industry or elsewhere, is a real challenge".

Why does it matter?

The EU has adopted hard regulation of AI in its AI Act – which is already in force. The UK has taken a lighter touch approach, relying on existing regulators to implement the UK's principles-based framework in their sectors in a context specific way. This approach was decided under the previous Conservative government.

The Labour government's election stance suggested a change in approach with an AI Bill expected to be published in the summer of this year to introduce binding regulation "on the handful of companies developing the most powerful AI models". The Bill was mooted to target genAI models like ChatGPT, to make existing voluntary commitments between companies and the government (such as those agreed at the UK's Safety Summit) legally binding, and to turn the UK’s AI Security Institute into an arm’s length government body.

However, during April's debate, in an apparent change in tack, Patrick Vallance told the Lords that the consultation due later this year would focus on newer advanced general intelligence and superintelligence models (ie not the genAI models that we are seeing at present).

He summarised the new cross-government AI approach as focusing on first of all:

existing regulators regulating their own areas, but in an even more joined up way and with more resource
the development of assurance tools, as outlined in the AI Opportunities Action Plan – this will result in the adoption of AI systems that have been assessed by properly validated assurance tools, and there will be a market in making sure that the validation system grows and becomes an important way of assuring users
the consultation (referred to above), will require "a cross cutting general piece of work", ie the current sector-by-sector approach might not be enough with a broader approach required for these powerful and adaptable AI models.

What action should you take?

Businesses will need to closely monitor the proposed consultation to see if there are yet more changes to the government proposals and plans. In the meantime, businesses should focus on their industry sector regulator for guidance, future plans and activity.

Key dates
UK government consultation "later" in 2025.

Helen Armstrong
Partner
+44 7912 519732
helen.armstrong@rpclegal.com

Ricky Cella
Senior Associate
+44 7895 301115
ricky.cella@rpclegal.com

Joshy Thomas
Knowledge Lawyer
+44 7927 578065
joshy.thomas@rpclegal.com

AI guide

Our AI guide is a user-friendly overview of some of the key legal implications of artificial intelligence, and the related issues that businesses will need to address when considering using artificial intelligence. In this AI guide, we have focused on the types of AI, and the related issues, that we see as most likely to be at the forefront for our clients in the short term. Our intent is for this to be a living document, and we will continue to add to the topics covered in this AI guide, so please do keep an eye on this page for updates.

Visit the AI guide