Is AI legislation coming to the UK?
The AI landscape is evolving rapidly. Accompanying this rapid pace of change is a growing awareness of the opportunities that AI presents for businesses, as well as increasing concerns over the potential risks surrounding the use of AI systems. The second global AI safety summit, co-hosted by the UK and South Korea, has just taken place in Seoul, building on the previous one hosted by Rishi Sunak in the UK in October of last year. To mitigate the risks that AI presents, many states have recently moved to some form of AI regulation.
While the UK has thus far taken a soft approach to legislating AI, it is clear that a change may be on the horizon, as demonstrated by the recent discussion of the UK Artificial Intelligence (Regulation) Bill in Parliament. Whilst this particular Bill has just been dropped with the dissolution of Parliament ahead of the next election, the possibility of AI legislation is far from over, and the next government will likely revisit the issue in due course. This raises interesting questions regarding the format that UK AI legislation might take, as well as the potential implications of this for UK businesses.
The UK’s approach to AI legislation so far
To date, the UK has not imposed any AI regulation on business. In March 2023, the UK government published its AI Regulation White Paper which argued that it was too early to legislate against AI, although this would be needed once understanding of the risks surrounding AI use had matured. The Paper proposed to split responsibility for AI supervision between existing UK regulators, providing the following cross-sectoral principles for regulators to interpret and apply within their respective sectors: safety, security and robustness, transparency and explainability, fairness, accountability and governance, and contestability and redress.
The White Paper (and subsequent Response published in February 2024) did not impose any statutory duties on these regulators, and noted instead that the progress of regulators and the evolution of AI risks would be monitored regularly. The White Paper and Response thus place the onus on regulators, rather than individual businesses, to monitor and mitigate AI risks. However, the notable support for and coverage of the recently fallen UK Artificial Intelligence Bill is a sign that some form of regulation regarding the development and use of AI by businesses may eventually be coming to the UK.
The UK Artificial Intelligence Bill
Introduced in November 2023 as a Private Members’ Bill to the House of Lords, the UK Artificial Intelligence Bill would have impacted UK businesses in two key ways. Firstly, it would have introduced the requirement on all businesses using AI to designate an AI officer. This officer would be responsible for ensuring the “safe, ethical, unbiased and non-discriminatory use of AI” throughout the business, and for ensuring that the data used by the business in any AI technology is “unbiased.”
Secondly, the Bill would have required companies to be audited by an independent third party accredited by the AI Authority body, which would be created by the act. Additional offences and requirements relating to the payment of fees, penalties and fines from businesses who fail to comply were also considered as part of the proposed regulation.
As mentioned above, Parliament has now been dissolved ahead of the general election and the Bill thus constitutes an unsuccessful first attempt at passing AI legislation. Nevertheless, it is a concrete step in a climate of growing discussion and awareness surrounding the impending necessity of regulating AI. Whilst short and high level in form, the Bill provides a window into the possible impacts of AI legislation on businesses.
Potential implications for UK businesses: insight from the EU AI Act
In preparation for potential regulation, UK businesses might look to the new EU AI Act for insights into the possible implications for businesses. While the Act takes a different approach to the proposed UK Bill, it nevertheless offers examples of the kinds of actions that might be required of UK companies, were UK legislation to be passed.
Approved by the EU Parliament in March 2024, the EU AI Act is considered to be the world’s first comprehensive horizontal legal AI framework. The Act adopts a risk-based approach, categorising AI systems according to four levels of risk, with differing requirements corresponding to each risk level.
Companies using AI systems defined by the Act as high risk (for example, systems that determine access to educational institutions or jobs) will need to comply with strict requirements including a risk-management system, human oversight, data governance, and controls ensuring cyber security, robustness and fairness. The Act also has an extraterritorial affect, applying to providers who place or put into service AI systems on the EU market, regardless of where they are located. Penalties for non-compliance are substantial and range between 1% and 7% of a company’s global annual turnover depending on the severity of the infringement.
All in all, the EU AI Act requires thorough due diligence from companies regarding their use of AI systems and marks a significant turning point in requiring businesses to adopt a proactive and risk-based approach. If similar requirements are adopted in future UK AI legislation, it would mark a watershed in how UK businesses must manage their use of AI systems.
The GoodCorporation view
There is still some way to go for UK AI legislation to be introduced, but it is clear that AI regulation should be on any company’s agenda if AI systems are used in their operations, as well as the government’s agenda. Some UK companies are already in scope of the EU AI Act, and the UK risks falling behind if it does not follow suit.
To prepare for potential UK legislation, companies should adopt a framework of ethical AI principles governing their AI use; monitor changes in AI regulation; invest in AI expertise; and train any employees using AI systems in their roles. Taking these initial steps will put companies in a strong position to respond to the introduction of any future AI legislation in Britain. It remains to be seen how the next government will pick up the baton on this issue.
To understand how GoodCorporation can help you in this area, please contact our team.