The UK's new AI Bill

The question

What will the UK's Artificial Intelligence Bill (the AI Bill) focus on and how will it regulate advanced generative-AI models?

The key takeaway

The recently announced AI Bill is expected to focus on the regulation of advanced AI models including generative-AI such as ChatGPT. The AI Bill also seeks to formalise voluntary agreements that companies had previously entered into under the former Conservative Government.

The background

The regulation of AI is a fast-moving area and the approach to regulation in the UK appears to be shifting. It was widely expected that an AI Bill would be announced during the King's Speech in July 2024, but instead the King simply noted that the Labour Government would seek to establish the appropriate legislation around powerful AI tools.

It has since been announced that an AI Bill will be introduced, and the regulatory approach taken by the new AI Bill is in contrast to the former Conservative Government's more laissez-faire approach, which was considered to be pro-innovation to encourage the growth of the AI industry, rather than pro-regulation.

The development

The UK's Secretary of State for Science, Innovation and Technology, Peter Kyle announced that the AI Bill, expected later this year, will focus on advanced AI models, including generative-AI tools such as ChatGPT. It is expected that the AI Bill will set out regulatory principles that apply to any business which develops, deploys, or uses AI.

The Labour government is seeking to formalise voluntary agreements that companies previously entered into under the Conservative Government. Under the former Government, a number of Big Tech companies including Amazon, Microsoft, Meta, and Google DeepMind signed voluntary commitments with the UK, US and Singapore governments. A notable voluntary commitment that tech companies signed up to in 2024 was in relation to risk, in that if severe risk could not be mitigated then companies would not deploy or develop the respective AI model at all (see here for our previous discussion on this topic).

Currently, if it becomes commercially beneficial to do so, companies can depart from voluntary commitments, which is an area of concern for AI regulation in the UK and has contributed to calls from senior UK government officials to formalise voluntary agreements and make these legally binding.

There will likely be an upcoming consultation on the new AI Bill before the end of 2024.

Why is this important?

As the first piece of primary legislation aimed at regulating AI in the UK, the AI Bill represents a departure from the current regulatory approach and changes the way certain advanced AI models are regulated in the UK. One big question will be how the AI Bill sits in comparison to the EU's AI Act which categorises AI technology based on its risk to society, requiring technologies with a greater risk of harm to comply with stricter regulations.

Any practical tips?

Companies that use or develop AI tools, including, generative-AI tools, should closely monitor the AI Bill as it progresses through Parliament and keep an eye out for any future developments that may affect their business. Additionally, companies that entered into voluntary commitments should consider the effects of the proposed formalisation on their use of AI tools.

Autumn 2024