The regulation of artificial intelligence (AI) has been a topic of discussion for both the UK and Northern Ireland. The British government wants to be at the forefront of AI innovation. However, Northern Ireland, influenced by its adherence to the European Union (EU) laws, will follow the EU’s stricter AI regulations. The EU was the first to publish sufficient regulations and those have become the global standard, with the delay of any AI specific legislation from the UK.
Though the delay is not without reason, as Sir Keir Starmer has said, the government will delay AI regulation to allow access to NHS data for research to boost growth. That being said, other areas of law such as GDPR apply to AI regulation in practice, and the prior framework remains relevant.
Now it looks like the previous government’s pro-innovation approach to AI regulation in the UK is going ahead. As demonstrated by the AI Opportunities Action Plan released this month, which we will be looking at in this blog.
The UK’s AI Strategy: Growth and Innovation
The government response to the AI Opportunities Action Plan was released last week. The government has accepted all 50 recommendations put forward by Matt Clifford. Keir Starmer has made clear he wants this plan to make the UK an AI superpower.
Within days of the election Clifford was commissioned to work on this plan. The goals of which are to boost the economy, improve public services including education and healthcare, and create personal opportunities for UK businesses.
The key priorities include:
Building world-class infrastructure
Investing in computing and data resources, expanding the UK AI Research Resource (AIRR), and establishing AI Growth Zones (AIGZs) like the first one at Culham.
Unlocking data assets
Leveraging public and private data for AI innovation, including initiatives like the National Data Library.
Developing AI talent
Attracting, training, and retaining top-tier AI researchers and entrepreneurs.
Safe and trusted AI development
Enhancing regulatory frameworks and supporting the AI Safety Institute to ensure responsible AI usage.
To plan for the UKs future AI needs a 10 year investment commitment which will be reached some time this year. By 2030 the UK’s AI Research Resource (AIRR) capacity will be expanded by at least 20x. The AIRR was established in 2023 to increase availability of GPUs for machine learning and AI researchers in the UK.
The Government will be establishing AI Growth Zones (AIGZ), areas with improved access to power. The first zone will be at Culham, headquarters of the UK’s Atomic Energy Authority (UKAEA) with agreement that the private-public partnership will benefit the local area. The government suggests the AIGZs have the potential to create job opportunities, improve local energy infrastructure and sustainability initiatives.
The Timeline
In the government’s response to the AI Opportunities Action Plan they have provided a delivery timeline for all of the recommendations the plan proposed.
Infrastructure
By Spring this year, the government will set out the majority of its infrastructure plans. DSIT will begin the process of selecting further AIGZs and looking into sustainability and the security challenges of AI infrastructure. Plans for allocation of compute will also be published as part of the long-term compute strategy. International compute partnerships to allow researchers to collaborate will also be set out in the long-term compute strategy coming this Spring.
The plan for a supercomputing facility capable of doubling the national AIRR will be set out later in the year.
Unlocking public & private data assets
In Spring, the government will be considering the role of government in international licensing of a copyright cleared British media asset training data set.
By summer this year, the government will be looking at how to use data assets in the public and private sector. The Department for Science, Innovation and Technology (DSIT) will be determining policy on how to use the National Data Library. Including its data access policy, concerning open government data sets for AI. As well as creating high-value data sets that will meet public sector, startup and academia requirements.
Also in DSITs coming data access policy will be how to incentivise and reward researchers and industry to curate and unlock private data sets.
Fostering AI Talent
The UK government is committed to growing its AI capabilities by focusing on developing a strong pipeline of skilled professionals.
By Spring the skills gaps will be assessed. DSIT, the Industrial Strategy Council, businesses, training partners and unions will work together with national and local governments, to assess the country’s needs. The Industrial Strategy will also put forward how the UK plans to attract skilled AI workers from abroad.
The majority of plans for fostering AI and digital skills in the UK will be published over the next two years. The government aims to increase the number of AI graduates by supporting universities to provide relevant degree courses and training. Aiming to improve gender balance, the government will encourage more girls to explore AI through extracurricular activities in schools.
To draw top AI talent to the UK, the government is exploring options like a prestigious AI scholarship programme. They are also looking to expand immigration routes to attract graduates from leading universities worldwide, with more Turing AI Fellowship programmes.
Safe and Trusted AI Development
All but one of the responses to the UK AI regulation recommendations will be published this Spring. The AI Safety Institute will continue to be supported and expanded to continue its research. DSIT will consult on proposed legislation and AISU will be established as a statutory body.
The government launched a consultation to reform the UK text and data mining TDM regime in December last year. Greater clarity is needed for the creative industry and AI developers to move forward.
Before the spending review the government will be examining the future needs of their regulators so they can fully drive growth and reduce AI risks.
DSIT will be working with regulators to accelerate AI in priority sectors with the most growth-potential, providing pro-innovation initiatives. Updates on this will be published Summer this year and then continuously thereafter. Regulators will also publicly report their AI innovation promoting activities. By Summer and continuously thereafter, DSIT will publish a progress update on regulators with the most AI activity.
DSIT, the Alan Turing Institute and UKRI will be working together to drive progress supporting the government’s five missions and attracting global talent. We can expect an update on this by Autumn this year.
By Spring 2026, DSIT will prioritise the funding for the expansion of the AISI’s AI safety fast grants programme. As well as DSIT’s own AI assurance tools, designed to assess if AI systems does what they’re supposed to do.
A Scan, Pilot, Scale Approach
The plan proposed a new approach to procurement to scale up where AI is being used well across the public sector. These instances are usually isolated and small scale, so the plan proposes DSIT support public sector partners to move faster. The majority of these measures will have updates available by Autumn this year.
Scan
By the Summer we can expect an update on DSITs exploring two-way partnerships with startups and AI vendors to signal future developments and public sector demand. A consistent framework for how to source AI will also be developed. So whether to build in-house or create challenges.
By Autumn and thereafter, DSIT will appoint AI leads to identify where AI solutions can be applied for each mission. DSIT will be building a cross government group with AI understanding that will do “technical horizon scanning and market intelligence”.
Pilot
All of the next measures will have updates in Autumn of this year.
As part of the new digital centre of government (which will be based in DSIT), a rapid prototyping capability will be created. Designed to build and test proof of concepts with in house AI experts and design and user experience specialists.
Currently DSIT is building on the work done to hire external AI talent by GDS, the Government Digital Service and i.AI, the Incubator for Artificial Intelligence (now a part of GDS as of Jan 2025). This would result in a technical senior civil servant stream to streamline recruitment.
Scale
DSIT is also working on developing i.AI’s experimentation environment to create a more streamlined approach to get access to datasets, large language models and infrastructure. DSITs AI procurement will also be getting improvements, with a faster multistage scaling procurement process. So that small-scale funding for pilots can be acquired faster.
A scaling service will also be set up for the successful pilots to ensure wide implementation. National AI tenders will be sorted by mission focus and DSIT will scope options to improve how AI is procured. DSIT also will be taking inspiration from Singapore’s GovTech, to develop or procure a tech stack that will support narrow and large language models for UK citizen interactions in the UK.
Lastly, interoperability, reusable, open source code will be mandated whenever appropriate, in line with the Technology Code of Practice.
Public & Private Cooperation
The UK government plans to actively advance collaboration between the public and private sectors to achieve its AI objectives.
By Summer 2025, DSIT will deliver several immediate measures to strengthen public-private collaboration and these include:
- Scaling 1-2 public sector-led AI pilots as open-source solutions.
- Launching a citizen-facing AI tool for personalised government engagement.
- Hosting hackathons aligned with national AI missions.
- Piloting the AI Knowledge Hub.
- Appointing AI leads for each mission to identify targeted solutions.
Smart procurement will be very important for the government going forward as it will be positioned as both a key customer and market shaper. As will how digital government infrastructure is used to support innovators. We can expect updates by Autumn 2025 on these.
The AI Knowledge Hub this Summer, will be designed to centralise best practices, case studies, and open-source solutions, so knowledge is available across sectors.
The government will also be creating a new function of government that will partner with AI companies and we can expect updates on this by the Spring.
Opportunities and Challenges
According to research commissioned by Microsoft AI could provide a £550 billion in economic value to the UK’s GDP by 2035. While the UK might have the most advanced tech sector in Europe, there are three areas which hold us back.
First, the UK doesn’t have the basic infrastructure to build data centres, a challenge that will only get bigger with the demand for further compute and data for larger AI models growing. Second, companies just aren’t aware of the benefits digital technology can provide them. Of the businesses that were polled, roughly half weren’t aware of any use cases. The last challenge is a shortage of digital skills for companies looking to incorporate AI.
AI Investment
Cultivating, recruiting, and maintaining future AI researchers and entrepreneurs while fostering secure and reliable AI progress and implementation through guidelines, safety measures, and guarantees.Finally, when suitable, the use of interoperable, reusable, and open-source code will be required, adhering to the Technology Code of Practice. Collaboration between public and private sectors.
The government’s drive to invest in sovereign compute is being reflected in the private sector. With more than £25 billion investment in new data centres in the UK so new AI technologies can be trained and deployed. So that the Uk can roll out AI faster in healthcare for example. Further details of the UK’s compute strategy will be published in spring.
A recent survey by the Boston Consulting Group (BCG) has demonstrated that 51% of businesses in the UK are planning to invest in technologies like AI, instead of hiring new staff.
Northern Ireland’s EU Alignment
Northern Ireland will follow a different path. Due to the Northern Ireland Protocol, the region must obey EU laws. This includes the new EU AI Act. Northern Ireland will follow the EU’s stricter AI regulations. For a bit more information about EU AI regulation have a look at the FAQs in our Overview of UK AI Regulations blog.
The differing rules could cause problems for businesses. Northern Ireland and the UK may face a ‘digital border’. This means businesses may need to follow two sets of rules. This could create confusion, especially for companies operating in both areas.
Looking Ahead: A Global AI Leader?
Despite the challenges, both the UK and Northern Ireland have strong plans for AI. The UK wants to be a global leader in AI and Northern Ireland will continue to follow the EU’s AI standards. Though this may lead to new opportunities for cross-border collaboration.
As both regions move forward, the impact of their different approaches will become clearer. The next few years will show how businesses adapt to the changing AI landscape in the UK and Northern Ireland.
