House of Commons Library publishes paper on AI and employment law

Paper provides an overview of the issues around AI and employment in the UK.

On 11 August the House of Commons Library published its paper: Artificial Intelligence and employment law.

The paper looks at the different types of AI and their current applications, noting that it is not new technology. Rather, a lot of existing technology has some, even if limited, background or “smart” processing that can be described as AI or AI-supported. However, increased interest in AI reflects recent developments that have significantly advanced automatic decision making, mimicking the process of giving an opinion or making a judgment on any issue.

Whilst ‘general AI’ (that is, an artificial intelligence that can be applied across any discipline and demonstrate an ability to learn in a way comparable to a human) appears to still be some way off, the paper provides a handy introduction to the key areas in which AI already interacts with employment, including the implications and interactions with common law, equality law, privacy and data protection law, highlighting the use of algorithmic management tools in recruitment, monitoring and surveillance.

For example, algorithmic management is prevalent in the ‘gig economy’ and used in app-based services like Uber and Lyft to monitor and manage their workers. This impacts on many aspects of the workers’ experience on the system, from allocating jobs to performance management and setting fares through to the feedback systems customers use to rate the drivers.

As referenced in the paper, a Department for Culture, Media and Sport report found that, as of January 2022, 68% of large companies and 15% of all UK businesses had adopted at least one form of AI and that a further 10% planned to adopt it. The main ways they had been used were through the use of spam filters, chat bots, facial recognition technology, voice assistants and personalised shopping.

According to data summarised in the paper, AI use was more prevalent in large companies, with the ICT, professional, scientific and technology sectors in particular leading the way in AI use. There are indications that employers see efficiency and cyber security as particular areas for future AI adoption.

The paper also provides an overview of the UK’s proposed way forward on AI regulation and the potential problems with this. In a 2023 White Paper, the UK established the key principles of AI regulation in the UK as safety, transparency, fairness, accountability and contestability.

The UK Government is expected to publish a UK AI regulatory roadmap at any point between now and Autumn 2024, but it remains to be seen whether this will happen or otherwise be affected by a likely general election in 2024. As an indicator of current intentions, the government has broadly set out its plans to use existing regulators to oversee AI in the UK, with the government expected to take responsibility for central coordination between the regulators.

Comment

Some have highlighted that the current proposals are relaxed and hands-off.

Concerns have been expressed over the plan to use existing regulators, with commentators pointing out that there are hundreds of regulatory bodies in the UK to coordinate on what is a new and potentially complex matter, with necessary funding so far lacking in current plans.

The European Commission on Human Rights has commented that current UK proposals provide insufficient safeguards against AI in respect of human rights. Other commentators have noted the UK’s approach diverges significantly from the EU’s impending AI Act which seeks to regulate the use and development of AI much more strictly, with clear parameters around what kinds of applications are permissible and not permissible. This could create problems for businesses operating across the UK/ EU.

On the other hand, some commentators have suggested the current approach signalled by the UK government is agile and pragmatic and more closely aligned with the present US approach, in that there is no clear guidance on AI application in the US so far.

2024 is therefore shaping up to be a significant year for AI development in the UK. At present, a general election is expected in 2024 and if Labour form the next UK government it seems the UK will move towards a more interventionalist approach on AI. This would appear to chime with Labour’s signalled intent to reestablish closer links with the EU which may lead to closer regulatory alignment under the recent ‘make Brexit work’ mantra.

In the meantime, all employers will face a steep learning curve on the applications, limits and complications of AI as it is integrated into working practices. Even where employers themselves have not directly implemented AI, it seems likely they and their staff already do, or are likely to see, the indirect application of AI by service providers.

Thought around the impact of this integration on privacy, personal data and fundamental aspects of the employment relationship including recruitment and line management will be needed if employers are to make the most of AI and avoid the associated risks.