23 September 2022

The UK Government recently published its AI policy paper, “Establishing a pro-innovation approach to regulating AI”. The paper sets out the policy the Government intends to pursue in the regulation of AI and adopts an interesting approach to the policing of a fast changing and constantly evolving sector.

As many will know, there is no specific regulation of AI in the United Kingdom. While some legislation indirectly relates to AI (for example, the Automated and Electric Vehicles Act 2018), there is currently no overarching framework or set of principles which govern the use of AI in the UK. The policy paper confirms that this model is set to continue, with no laws proposed specifically to regulate AI.

That is not to say that AI is unregulated at all, for example, intellectual property and data protection legislation go some way to regulating ownership of AI and AI output and how personal data used in connection with AI must be processed. Nonetheless, this legislation is not specific to AI and still gives rise to queries which arise only in the context of AI, as demonstrated by Dr Stephen Thaler’s application to the Court of Appeal to allow registration of a patent in the name of his AI machine “DABUS”.

It is clear from the policy paper that the UK Government intends to create a broad regulatory framework which sets out principles which should be observed in the governance of AI. It is intended that detailed governance will fall to sector regulators, allowing them the opportunity to address specific industry issues.

The emphasis of the paper is on innovation and creating an environment in which AI can flourish, allowing developers, inventors and businesses the degree of creative freedom to lead in their fields of interest.

The policy paper identifies a number of cross-sectoral principles which will apply to AI regulation, namely:

1. Ensure that AI is used safely
2. Ensure that AI is technically secure and functions as designed
3. Make sure that AI is appropriately transparent and explainable
4. Embed considerations of fairness into AI
5. Define legal persons’ responsibility for AI governance
6. Clarify routes to redress or contestability

The principles will, at least initially, not be statutory in nature, but will instead act as guidelines for industry regulators who will have some degree of flexibility and freedom in determining how to police use of AI in their sector.

Continuing the theme of the UK breaking away from the EU, this light touch approach to regulation is in stark contrast to the EU position where harmonised rules for cross-sectoral regulation is the order of the day.

The UK Government’s approach provides the opportunity to anticipate changes and respond quickly to industry issues, allowing for governance which reflects practical application of AI. However, it also gives rise to the risk of piecemeal and potentially conflicting regulation, particularly where the AI in question spans a number of sectors and involves more than one regulator. The possibility of regulation developing in different directions in different sectors is something that may yet require centralised oversight.

The policy paper does not set out the detail of how this approach will be implemented with a white paper due to be issued later in the year.

Lynn Richmond, Partner & Accredited Specialist in Intellectual Property: lyr@bto.co.uk / 0131 222 2939