The UK's AI action plan

Posted: 23/09/2022

AI Paper

The UK government published its policy paper on the future regulation of artificial intelligence (AI) on 18 July 2022 (the AI Paper), taking a less centralised approach than the EU. This follows the National AI Strategy which was published in September 2021.

Ahead of the White Paper which is due to be published in 2022, the AI Paper provides assistance for businesses to prepare for the upcoming regulation of AI in the UK. AI regulation has been a hot topic of debate in the past year and there is a general consensus that regulation is required but the debate has surrounded the mechanisms for doing so.

The European Commission has recently published a draft regulation that has a risk-based approach. Some systems have been banned and others are strictly regulated. The regulation has also introduced fines and national supervisory authorities which some have noted may affect innovation and impose onerous obligations on businesses. The AI Paper has attempted to address the approach that the UK may take in regulating AI, including how it may differ from the European approach.

Overview

Under the AI Paper, it will fall to individual regulators such as the Financial Conduct Authority (FCA) and the Competition and Markets Authority (CMA) to interpret and implement the principles (described as the context-based approach). The regulators will be encouraged to consider ‘lighter-touch options’ that could include guidance and voluntary measures or by creating ‘sandboxes’ – ie trial environments in which businesses can examine AI tech before it is introduced to the open market.

Although the AI Paper has not defined AI, it has identified two characteristics in order to set the scope of the UK regulation. These are the adaptiveness of technology and autonomy from humans. The adaptiveness of technology characteristic relates to the fact that AI systems often partially operate on instructions which have not been expressly programmed with human intent and are often trained and execute patterns which are not easily discernible to humans. The second key characteristic, autonomy of the technology, relates to the fact that decisions can be made without the express intent or ongoing control of a human.

Digital Minister, Damian Collins, explained that an aim of the AI Paper is to “make sure the UK has the right rules to empower businesses and protect people as AI and the use of data keeps changing the ways we live and work”. The way in which the AI Paper has been drafted allows the government to capture forms of AI which are not yet in existence and allow individual regulators to develop more detailed definitions according to their relevant areas.

Pro-innovation approach

The AI Paper differs from the European approach by taking a pro-innovation approach. It is context-specific and has been designed to allow regulators to focus on the genuine and harmful risks to fostering innovation. Another key area highlighted was proportionality that would allow regulators to consider lighter enforcement. This pro-innovation approach contrasts with the European approach as it focuses on allowing innovation and provides for less onerous enforcement. Context is a key consideration for the AI Paper and this provides for less uniform regulation.

Cross-sectoral principles

In order to provide some consistency across different sectors, the AI Paper has provided for cross-sectoral principles which will be applicable to all regulators. The principles are as follows:

• Ensure that AI is used safely – it is important for all regulators to take a context-based approach to assessing the risks in their area.
• Make sure that AI is technically secure and functions as designed – systems should be technically secure and do what is intended and claimed in order to ensure public confidence in the functioning of the AI systems.
• Make AI appropriately transparent and explainable so that decision-making is easily understood by users.
• Embed considerations of fairness into AI – outcomes should be justifiable.
• Define legal persons’ responsibility for AI governance – accountability for outcomes produced by AI and legal liability must rest with an identifiable legal persons such as a real person or body corporate.
• Clarify routes to redress or contestability – there should be clear measures to contest outcomes so that the use of AI does not remove an affected group’s ability to contest an outcome.

As is consistent with the approach taken in the rest of the AI Paper, the principles are over-arching and high level.

Next steps

The UK government has invited firms to respond to a number of specified questions by 26 September 2022, with the forthcoming White Paper and public consultation expected in late 2022. The White Paper will consider the practical approach and is intended to consider whether it adequately addresses AI specific risks, the role and powers of the various regulators, the need for co-ordination and methods for monitoring, and evaluating the framework over time.

It will be interesting to see the response from firms to the proposals contained in the AI Paper and whether the lighter touch approach is favoured in this regard. The differences between the European and UK approach give rise to the question of how these two approaches will interact with each other.

The commercial dispute resolution team at Penningtons Manches Cooper can assist with any queries in relation to the proposed AI Paper and related issues. Please contact Laura Coleclough at laura.coleclough@penningtonslaw.com.

This article was co-written with Kate Abercromby, trainee solicitor. 

Return to news headlines