March 17, 2023 - Milton Keynes, England
AI Regulation: Developments in the UK's approach
by Shoosmiths LLP
Amongst the pensions, childcare and other headline financial announcements in this week's Spring Budget, Jeremy Hunt signalled the Government's continued focus on Artificial intelligence (AI) as a targeted growth sector for the UK economy.
The Chancellor pledged a £900m investment to support the development of supercomputing capabilities that can underpin the growth of the AI sector in the UK.
The announcement came at the same time as publication of Sir Patrick Vallance's wide-ranging review looking at pro-innovation approaches to regulation of technology, including AI. The review encourages a "step-change in the UK's regulatory approach to AI".
On one level this builds on the UK Government's stated approach to AI regulation up to this point, which leans heavily on the opportunity for post-Brexit divergence from equivalent regimes being developed elsewhere, most obviously the 'full fat' regulatory approach taken in the AI Act currently being finalised by our neighbours in the EU.
In other respects, though, the review challenges some of the Government's previous thinking on how regulation could be managed across different sectors and industry regulators. That sector-specific approach has always raised questions as to how companies developing AI solutions with general application, or application across multiple regulated sectors, could plan and invest in AI solutions with confidence. Sir Patrick's review calls that out, pointing out that whilst the risks of AI do vary depending on sector context, this divergence of approach could "further compound an overall lack of regulatory clarity".
Recommendations for AI in the review include:
- the rollout of a regulatory sandbox for AI products and services, to support industry innovation and encourage greater co-operation between regulators;
- a review and reform of existing intellectual property (IP) law to address the impacts of generative AI solutions like ChatGPT, which rely on access to huge data sets, text and images to generate new content; and
- a linked review of the ability for industry to access public data, and changes to support wider data sharing across the public sector.
Behind each of these recommendations there remain complex questions as to how existing laws, particularly in relation to IP and data protection, can be reconciled with the agile and evolving nature of AI technologies. What the review, and the Chancellor's investment in AI, do point to however is that the UK remains determined to plot its own path on AI regulation.
Read full article at: https://www.shoosmiths.co.uk/insights/comment/ai-regulation-developments-in-the-uks-approach