While the focus of attention in the world of AI has been the EU AI Act: EU AI Act Agreed – Discerning Data in recent weeks, there have also been some other noteworthy legislative developments. On 22 November 2023, the Artificial Intelligence (Regulation) Bill (the “Bill”) was introduced to the UK Parliament and passed the first reading in the House of Lords. The Bill seeks to establish a central AI authority (“AI Authority”) to oversee the UK’s regulatory approach to AI. The proposal for an AI Authority comes after the UK Government formally announced a UK AI Safety Institute at the global AI Safety Summit at Bletchley Park (summarised here).
Whilst the Bill largely reflects the approach of the UK Government, this is a Private Members’ Bill (“PMB”). PMBs are legislative proposals introduced into one of the UK Houses of Parliament by ‘backbench’ members (members who are not Government Ministers). Most PMBs fail to pass unless the UK Government steps in to support their progress through the legislative process.
Nevertheless, the Bill still provides a useful insight into how the UK Government could legislate in this area, . The Bill is light on detail, in keeping with the UK Government’s ‘pro-innovation’, light-touch framework for regulating AI and somewhat aspirational. Notably, however, the Bill provides no effective mechanisms to monitor compliance with its rules or to sanction non-compliance by businesses, although it does give the Secretary of State discretion to create offences with related penalties, fees and fines.
Definition of AI
The Bill defines ‘AI’ as technology enabling the programming or training of a device or software to: perceive environments through the use of data; interpret data using automated processing designed to approximate cognitive abilities; and make recommendations, predictions or decisions, with a view to achieving a specific objective. ‘AI’ includes generative AI, meaning deep or large language models able to generate text and other content based on the data on which they were trained.
The Bill proposes that the new AI Authority will have the following functions:
- ensuring that relevant regulators take account of AI;
- ensuring alignment of approach across relevant regulators in respect of AI;
- undertaking a gap analysis of regulatory responsibilities in respect of AI;
- coordinating a review of relevant legislation, including product safety, privacy and consumer protection, to assess its suitability to address the challenges and opportunities presented by AI;
- monitoring and evaluating the overall regulatory framework’s effectiveness and the implementation of the regulatory principles outlined below);
- assessing and monitoring risks across the economy arising from AI;
- horizon-scanning, including consulting the AI industry, to inform a coherent response to emerging AI technology trends;
- supporting testbeds and sandbox initiatives to help AI innovators get new technologies to market;
- accrediting independent AI auditors;
- providing education and awareness to give clarity to businesses and to empower individuals to express views as part of the iteration of the framework; and
- promoting interoperability with international regulatory frameworks.
The Bill proposes that the AI Authority must have regard to the following principles when regulating AI:
- safety, security and robustness;
- appropriate transparency and explainability;
- accountability and governance; and
- contestability and redress.
The Bill also includes requirements on the AI Authority to ensure that businesses which develop, deploy or use AI are transparent about their use of AI, test the AI thoroughly, and comply with applicable laws, including in relation to data protection, privacy and intellectual property. Further, the Bill proposes that the AI Authority will need to ensure that AI and its applications comply with equalities legislation and is designed in such a way that it is inclusive by design, and does not discriminate unlawfully among individuals or perpetuate unlawful discrimination arising in input data. There are also proposed obligations to ensure that AI meets the needs of those from lower socio-economic groups, older people and disabled people, and that it generates data that is accessible, interoperable and reusable.
The Bill proposes that any restrictions or burdens in respect of AI should be proportionate to the benefits, taking into consideration the nature of the service or product being delivered, the risk to consumers and others, the cost of implementation, and whether the burden or restriction enhances UK international competitiveness.
Transparency, IP Obligations and Labelling
The Bill also proposes measures requiring the Secretary of State to make regulations requiring any persons involved in training AI to supply to the AI Authority a record of all third-party data and intellectual property used in that training, and to assure the AI Authority that all such data is used with informed consent and complies with applicable intellectual property and copyright obligations. Suppliers of products or services involving AI would be required to give customers clear and unambiguous health warnings and labelling and opportunities to give or withhold ‘informed consent’ in advance. Any business which develops, deploys or uses AI would have to allow independent third parties accredited by the AI Authority to audit its processes and systems.
As a PMB, the Bill is unlikely to pass into legislation. However, although its position on the Bill is not yet clear, the UK Government has signalled an intent to legislate in this area. Therefore, the Bill gives an indication of what such legislation might look like although such provisions would inevitably be amended and diluted during the legislative process.
A date for the Bill’s second reading in the House of Lords is yet to be announced. We will continue to monitor developments in this area.