Beyond the AI Safety Summit Immediate priorities for UK Government

Six projects are awarded grants from round three of the UK Government's Levelling Up Fund to support major regeneration and transport projects.

The first ever Global AI Safety Summit, which took place earlier this month in Bletchley Park, marked an important moment in the development of AI. We applaud the UK for taking the initiative to lead from the front on this issue and bring together global leaders from across government, industry, civil society and academia, to better understand and mitigate the potential risks associated with Frontier AI.

With the successful Summit now behind us, UK Government should build on the momentum of the last few weeks and focus on delivering a clear approach to AI governance; taking into consideration the discussions that took place during the Summit and pre-Summit engagement events. Crucially there is a need to focus on the so called ‘Here and Now' issues as well as the frontier developments which were the focus of the Summit.

Given the scale and pace at which this technology is moving, and since many UK regulators are already pressing ahead with their own work in this area since the White Paper was published back in March 2023, there is an urgent need for Government to pick up the pace on their plans for AI governance. This is essential for providing industry with the clarity and certainty they need to invest here in the UK as well as helping to build public trust and confidence in the use of the technology.

techUK would therefore like to see Government prioritise the following areas as a matter of urgency before the end of 2023:

1. Publish a clear roadmap for implementation of the UK AI White Paper as part of Government's response to the AI White Paper consultation. This needs to be a clear action plan with priority areas mapped to a clear timeline. Particular focus should be given to the process by which regulators will translate the AI principles highlighted in the White Paper into practice, as well as details on how different AI applications will be deemed “high risk”. The roadmap should also include a detailed engagement plan setting out how industry and other key stakeholders will be consulted on next steps.

2. Provide further clarity on how the new ‘Central Function' will operate in practice. This includes sufficient resources, in the form of expertise and funding, to get this new body up and running now. This new Central Function is an essential component to the UK's plans to govern AI and will be integral to its success. Run effectively, this body will help to mitigate the risk of duplication, overlap or contradictory governance measures coming out of different regulators. In addition, the Central Function should have the resources to carry out an effective horizon scanning function to anticipate and plan for how this technology may evolve in the future, whilst also having clear engagement channels to industry, civil society and academia to effectively address any arising challenges.

3. As part of the upcoming Autumn Statement, secure sufficient funding for regulators to build their capacity and capabilities to fulfil their new roles and responsibilities, as highlighted in box 3.1 of the White Paper. This includes implementing the AI principles, determining high risk applications within their remit, providing supplementary guidance and monitoring/reacting to the latest AI developments in their field, to name but a few. To carry out these roles to a high standard many of the +90 UK regulators will need additional support in the form of funding and expertise. This funding is vital now to ensure a coordinated and joined up approach across the UK regulatory landscape.

Looking ahead to 2024 however, it is not only how we get AI governance right in the UK that is important. The other area we need to focus on is AI adoption. We need to step-up efforts to support the long tail of British businesses who currently don't have the ability to access the benefits of AI and risk being left behind. This involves refocusing on some of the critical enabling actions that will help to support broader adoption of AI, including access to skills and talent, compute, high quality data sets and data infrastructure. As well as showcasing some of the commercial benefits that come from AI adoption through case studies and providing access to advice and support on how to adopt AI. We are currently unaware of an advice hub that caters for these needs and requirements across government.

The AI Safety Summit provided a brilliant platform for highlighting some of the most significant opportunities and potential risks associated with future developments in AI. Whilst we must continue to have these important conversations to anticipate potential long-term, existential risks, we cannot afford to lose sight of the near-term, real-world challenges people and industry face today. By focusing on getting the UK's approach to AI governance right, government will be helping to address some of the more immediate challenges, and at the same time help to positively shape the long-term trajectory of this technology and how it's used in the future. There has therefore never been a more important time to move forward with pace and urgency to establish a clear and focused action plan on AI governance to support responsible AI innovation and adoption in the UK, and as we move forward with these plans, techUK stands ready and willing to help.

How do we ensure the responsible and safe use of powerful new technologies?