The AI Opportunities Action Plan
15 min read
The AI Opportunities Action Plan sets out how the UK will build, use, and benefit from artificial intelligence. It's a strategy to ensure the country doesn't just adopt AI tools created elsewhere, but develops its own capabilities and shapes how the technology evolves. If you work in the public sector, this plan will shape the AI tools available to you, the training you receive, and how your organisation approaches AI adoption.
Why the plan exists
The plan was commissioned in July 2024, shortly after the general election. Matt Clifford CBE, tech entrepreneur and Chair of the Advanced Research and Invention Agency, wrote it. The government published it in January 2025 alongside its formal response.
Britain is the world's third-largest AI market. Companies like Google DeepMind, ARM, and Wayve are based here. UK universities produce world-class AI research. The AI Security Institute has established global leadership on AI governance.
But the government recognised this foundation isn't enough. The US and China are investing heavily in AI development. Other countries are moving fast to build infrastructure and attract talent. Without a clear strategy, the UK risks becoming a user of AI technology rather than a creator of it.
The plan addresses this. It focuses on three areas. Building the infrastructure AI needs. Using AI to improve public services and boost productivity. Securing the UK's position in developing frontier AI capabilities.
The three pillars
Lay the foundations. Build the compute capacity, data access, skills base, and regulatory environment that AI development requires.
Change lives by embracing AI. Deploy AI across government and the private sector to improve services and drive economic growth.
Secure the UK's future with homegrown AI. Ensure the UK has companies at the frontier of AI development, not just as consumers of technology built elsewhere.
Each pillar contains specific recommendations. The government accepted all 50 actions and set timelines for delivery. An AI Opportunities Unit in the Department for Science, Innovation and Technology tracks progress.
Building the foundations
Compute infrastructure. AI requires massive processing power. The plan calls for expanding the UK's AI Research Resource by at least 20 times by 2030. This gives researchers and startups the capacity to train AI models.
The government committed to doubling AIRR capacity through a new supercomputing facility. Researchers and small businesses gained access in early 2025. Isambard-AI, a supercomputer at Bristol University, launched in July 2025. £2 billion was committed to expand UK compute capacity.
AI Growth Zones were created. These are areas with streamlined planning and energy access to accelerate data centre buildout. Five zones have been designated across Great Britain. They've generated £28.2 billion in investment and created over 15,000 jobs. The first zone is at Culham, headquarters of the UK Atomic Energy Authority.
Data access. AI systems need high-quality data to train on. The National Data Library was established to unlock public sector datasets responsibly. It received over £100 million at Spending Review 2025.
The Health Data Research Service launched with up to £600 million in funding. It provides secure access to health datasets for researchers. An AI Education Content Store is being developed to give AI systems access to curriculum material.
The government also committed to resolving copyright uncertainty. Under current UK law, using large datasets of copyrighted material to train AI models can infringe copyright. In December 2024, the government published a consultation proposing an opt-out scheme. Over 11,000 people responded. 88% said AI developers should be required to obtain licences for copyrighted material. Only 3% supported the opt-out proposal. The government has shifted to a more consultative process. The Data (Use and Access) Act 2025 requires the Secretary of State to produce an economic impact assessment and policy proposals by 18 March 2026.
Skills and talent. The plan calls for training tens of thousands of AI professionals by 2030. The AI Skills Hub launched in January 2026 with the AI Skills Boost programme. Every adult in the UK can access free AI courses that meet Skills England's foundation skills for work benchmark. Over one million courses have been completed since the programme began. The government aims to upskill 10 million workers by 2030.
New scholarship programmes launched. The Spärck AI scholarship supports master's students at leading universities. TechGrad undergraduate scholarships provide work placements in AI industries. The Global Talent Taskforce doubled its resources to attract international AI talent.
Regulation and safety. The plan maintains the UK's sector-specific approach to regulation rather than creating a single AI law. The AI Security Institute received £240 million at Spending Review 2025. It has tested 30 frontier models and published research on AI capabilities and risks.
Regulators are being funded to build AI expertise. They're required to publish annually on how they've enabled AI innovation in their sectors. Regulatory sandboxes allow companies to test AI applications in controlled environments with reduced regulatory burden.
Adopting AI across government and business
The plan proposes a "scan, pilot, scale" approach for government.
Scan. Identify where AI could solve problems. Appoint AI leads for each government mission. Build teams that understand both technology and specific challenges. Create partnerships with AI companies to understand upcoming developments.
Pilot. Test solutions quickly. The plan recommends faster procurement processes that don't disadvantage startups. Small-scale pilots should get easy access to funding. Bureaucratic controls only layer on as investment grows.
A rapid prototyping capability was created. Data-rich experimentation environments give teams access to datasets, language models, and compute. Specific support helps hire external AI talent.
Scale. Roll out what works nationally. The plan recommends a dedicated scaling service with central funding. Successful pilots should expand beyond departmental boundaries.
The NHS AI Diagnostic Fund shows what this looks like. £21 million was allocated to 12 imaging networks covering 66 NHS trusts. One-third of NHS chest X-rays are now AI-assisted. That's 2.4 million scans.
Other examples in development include AI tools to reduce planning processing times, AI-powered meeting scribes for local authorities, and trials of AI tutoring tools to support up to 450,000 children on free school meals.
Business adoption. The Modern Industrial Strategy, published in June 2025, sets out how AI can drive productivity across eight priority sectors. AI Sector Champions were appointed for each. They work with industry and government to identify high-value AI applications and overcome adoption barriers.
£150 million was committed for AI and technology programmes. Innovate UK's BridgeAI programme expanded to provide guidance and funding to thousands of businesses.
An AI Knowledge Hub was launched in May 2025. It publishes tools, prompts, how-tos, best practices, case studies, and open-source solutions shared by the public sector community. Organisations can find resources organised by capability such as finding information, analysing data, creating documents, and running meetings.
Securing homegrown AI capabilities
To secure the UK's position in frontier AI, the plan recommended creating the Sovereign AI Unit. This unit partners with private sector AI companies to ensure the UK has national champions at the frontier of AI development.
The unit was established with up to £500 million in funding. It invests in UK AI companies working on critical parts of the AI value chain. It also provides non-financial support. Access to compute through AI Growth Zones. High-value datasets. Help attracting international talent. Collaboration with the national security community.
Early investments include £8 million for the OpenBind consortium's structural dataset to unlock AI-driven drug discovery. The unit allocated sovereign compute to support researchers and startups, including the University of Cambridge's MACE materials foundation model.
The rationale is clear. Leading AI companies enjoy significant advantages. New competitors are unlikely to emerge without strategic support. Countries that develop frontier AI will have outsized influence over how the technology evolves and major economic benefits.
The plan describes this as an "asymmetric bet." The potential upside from success far outweighs the investment required. The cost of inaction isn't just economic. It risks leaving the UK dependent on foreign AI systems with little say in how one of the century's most powerful technologies develops.
What this means for public sector workers
These developments affect your work in several ways.
You'll use AI tools. The government is scaling successful pilots nationally. Tools that reduce administrative burden, speed up processing, or improve service delivery will roll out more widely.
Training will be available. Free AI courses are available now. More structured programmes are launching through Skills England. The aim is to ensure workers can use AI effectively and adapt as the technology evolves.
Procurement will change. Faster, less bureaucratic processes are being developed for AI tools. The goal is to make it easier to test solutions quickly and scale what works.
Your sector may get specific support. AI Sector Champions are developing adoption plans for priority industries. Regulatory sandboxes may make it easier to test AI applications in your area.
One year on
By January 2026, the government reported meeting 38 of the 50 actions in the plan.
Key milestones include five AI Growth Zones designated, Isambard-AI supercomputer launched, National Data Library funded and operational, over one million people completing AI training courses, AI Security Institute testing 30 frontier models, and multiple AI tools deployed in public services.
A public dashboard tracks progress against all 50 actions. This shows which commitments have been met, which are in progress, and timelines for delivery.
What happens next
The government published the UK Compute Roadmap in July 2025. This sets out infrastructure needs through 2030. Up to £2 billion will expand the AI Research Resource twentyfold. £750 million will fund a new national supercomputer at Edinburgh, expected in 2027. The roadmap identifies that the UK needs at least 6GW of AI-capable data centre capacity by 2030.
The AI Growth Zone Delivery Unit is working to build out the first wave of zones. The government aims to facilitate buildout of cutting-edge AI infrastructure that delivers both national and local benefits. Additional zones will be designated through a selection process.
The Sovereign AI Unit continues investing in UK AI companies. Its next phase launches in April 2026. The unit will deploy up to £500 million in funding to support companies working on critical parts of the AI value chain.
Public service AI tools are moving from pilot to national deployment. The government aims to make the UK the fastest AI-adopting country in the G7. Tools that have proven successful in trials will scale across departments and local authorities.
The AI and Future of Work Unit was announced in January 2026. Backed by an expert panel from business, academia, and trade unions, it will research and monitor AI's economic and labour market impacts. It will provide timely advice on when new policies should be implemented across government.
Skills programmes are expanding. The AI Skills Boost programme aims to reach 10 million workers this decade. TechLocal funding of £27 million will help create up to 1,000 tech jobs in local communities and enable new AI degree programmes and graduate traineeships.
How this fits with other AI policy
The AI Opportunities Action Plan focuses on seizing economic and public service benefits. It sits alongside other government AI initiatives.
The AI Regulation White Paper set out principles for governance. The Data (Use and Access) Act 2025 created requirements for reporting on AI and copyright. The Cyber Security and Resilience Bill addresses AI-critical infrastructure.
Together, these create a framework where AI development is supported, adoption is accelerated, and risks are managed.