What did the government say in its AI action plan launch?

The Labour government, which is finding economic growth elusive, has turned to artificial intelligence (AI) to help turn around the country’s fortunes and wants to make the UK one of the top three global centres for the technology.

In January the prime minister, Keir Starmer, promised that AI will be ‘mainlined into the veins’ of the country and receive investment running into the billions as part of a 50-point action plan intended to make the UK an ‘AI maker’ not an‘ AI taker’. The other key goals of the AI action plan are to ‘push hard on cross-economy AI adoption’ and to invest in the foundations of AI. 

These are the main areas the government will focus on in the AI Opportunities Action Plan.

1. Homegrown AI

Ministers want the amount of computing power (known as ‘compute’) under public control to increase by 20 times by 2030. Some of this capacity will be used to create UK AI models powered by a new supercomputer, potentially creating rivals to the likes of ChatGPT. To do this there will need to be a investment in data gathering and storage, which will take the form of a national data library. The plan is for there to be five major datasets, one of them probably NHS patient data, which has substantial commercial and research value. The government wants to go further than just make available the datasets it has – rather it wants to take a more strategic approach to data by shaping it to maximise its utility and value.

2. Private sector involvement

AI growth zones will be set up around the country. Oxford and Cambridge are candidates for this, given their academic, commercial and research credentials. The action plan envisages the creation of an AI headhunting unit and changing visa rules to bring in talented people from overseas. It proposes the creation of ‘sector champions’ who would encourage the use of AI within their industries, which could be anything from the media and creative industries to banking.

3. AI and the public sector

The plan is for public services to be delivered with the help of AI. To start with,there will be a range of pilots, the best of which will be used to create nationwide services. The hope is that this will improve public sector productivity and service levels. There will be a focus on automating routine tasks to save money.  Examples of potential uses include helping teachers prepare lessons and deploying AI-powered cameras to report potholes. 

4. AI and energy generation

AI computers need a lot of power so the AI growth zones will need extra grid capacity to be built. The government says it will set up an AI energy council to support the expansion of green energy and seek investment for new small nuclear reactors and renewable energy generation.

5. AI and copyright

Unpopular changes are afoot here. The government wants to change copyright laws so that rights holders will have to take the initiative to stop their work being scooped up by AI firms and used in their databases. Currently, copyright holders have to give consent for their work being used in AI datasets even though in practice it has not been sought by companies in the US and China. The government says its approach is in line with the EU’s and is necessary to maintain competitiveness. 

6. Skills

New technology requires new skills. The government will assess the gap between the  skills UK workers have and the ones that they need. It will then put more resources into AI training in schools and colleges, using the Lifetime Skills Guarantee programme to support workers with knowledge and skills development throughout their careers. 

The government concedes it does not know what the AI-skilled workforce of the future will look like and recommends looking at countries such as South Korea and Singapore that have already made substantial moves in AI to provide insights.  

AI and data security

Some elements of the AI Opportunities Action Plan touch on the issue of data security. Healthcare records, in particular, are highly sensitive and the government is mindful of the need for public confidence in the security of the datasets it proposes creating. It says it will explore synthetic data generation techniques to construct versions of the most sensitive datasets to protect people’s privacy. 

However,developers will need access to high-quality data, which means opening up public and privately held datasets to them. It is simply too early to say what the regulatory environment will be that AI firms will have to operate in; however, there will inevitably be tensions between access to data and maintaining individuals’ privacy. Furthermore, there are many unanswered questions, in particular about people’s ability to opt out of their data being included and which companies will have access to their data.  What cybersecurity measures will be in place is another concern, especially when it comes to datasets used in critical infrastructure, healthcare and education.

Another problem is explaining to people how their data is being used, given AI often uses overlapping databases and the ‘black box’ nature of some of its processes,whereby the steps it takes to reach a decision or action are occluded from view.

AI and energy security

The UK’s power generation and distribution capacity is going to rely more and more on AI. Its main roles will be to improve forecasting and optimisation of renewable energy generation, improve energy load forecasting, predict maintenance needs, optimise energy storage and support cybersecurity. It will also be needed to design and operate an increasingly complex energy market,composed of unpredictable wind and solar generation, peer-to-peer energy trading, small scale feed-ins to the grid as well as base load from nuclear and gas.

AI’s ability to significantly improve efficiency through these optimisations will enhance UK energy security by enabling it to maximise its resources and operate the more fragmented energy systems of the future, thus making the country more self-sufficient. Given the unpredictability of renewable generation it will only grow in importance as energy inputs to the grid move away from gas.

However, data privacy is a concern. Energy companies have access to large amounts of personal, real-time data from smart metering, which could be useful to hackers and other criminal enterprises.

What else is the government planning this year?

At the end of January the chancellor, Rachel Reeves, unveiled the Oxford-Cambridge growth corridor, saying it has the potential to become Europe’s Silicon Valley. No doubt AI will play an important role in this.

Later on, we can expect publication of the digital and technologies sector plan, the compute strategy, which will be unveiled in the spring, followed by the digital and AI road map in the summer.

Finsbury Robinson is a full-service tax, accountancy and business advisory firm. Our friendly and highly experienced team is available on 020 8858 4303 or via email at info@finsburyrobinson.co.uk