Government Backs British Steel, AI and Shipbuilding for National Security Contracts

The UK Government has announced that steel, shipbuilding, artificial intelligence and energy infrastructure will be formally recognised as sectors critical to national security. New guidance published on 26 March 2026 instructs government departments to prioritise contracts with British businesses in these four areas. The announcement also forms part of a much larger government programme to invest heavily in AI infrastructure involving billions of pounds in data centre spending, electricity subsidies, and large scale software partnerships.

What Has the Government Announced on Procurement?

The Cabinet Office has published new guidance telling government departments how to apply national security considerations when buying goods and services. For the first time, four sectors are explicitly named as strategically critical, steel, shipbuilding, artificial intelligence, and energy infrastructure.

In practice, this means that when departments are awarding contracts, they will be expected to consider British suppliers in these sectors as a priority. For steel in particular, departments will be required to either use British made steel or provide a formal written justification if they intend to source from overseas. This requirement builds directly on the UK Steel Strategy, which was launched the previous week.

Cabinet Office Minister Chris Ward said:

"This Government is backing British businesses and the working people who power them. These reforms are about using the full weight of Government spending to support British jobs, protect our national security and grow our economy. Whether you make steel in Scunthorpe, build ships on the Clyde or run a small tech firm in the Midlands, this Government is on your side."

The Public Interest Test

Alongside the national security rules, the Government has introduced a new Public Interest Test. Under this test, departments will now be required to assess whether outsourced service contracts worth more than £1 million could be delivered more effectively by public sector workers instead. The Government says this test will apply to over 95% of central government contracts by value. The stated aim is to end what ministers describe as "outsourcing by default", the longstanding practice of automatically contracting out services without properly evaluating whether doing so represents good value.

Social Value in Contracts

The Government also announced that companies bidding for contracts will be encouraged to include commitments to local jobs, apprenticeships, and community benefits as part of their bids. For contracts over £5 million which the Government says covers more than 90% of central government contracts by value, departments will be required to publish and report annually on a specific social value goal. The aim is to use government spending not just to buy a product or service, but to strengthen local economies and skills pipelines.

AI Tools for Procurement

The Government has also announced the development of a suite of AI tools to streamline the procurement process. Contract terms will be simplified, and business information will be integrated into a central platform so that small businesses do not have to submit the same information multiple times when bidding for different government contracts.

Why Is the Government Doing This?

The Government's stated reasoning is that recent global events have exposed the fragility of international supply chains. Ministers argue that the UK cannot rely on overseas suppliers for goods and services that are essential to national security. This logic applies especially to sectors where domestic capacity has declined significantly over recent decades.

The procurement changes are designed to align with the Government's broader Modern Industrial Strategy, which aims to provide long term investment certainty in strategic sectors, and with the National Security Strategy, which the Government says aligns national security objectives and economic growth plans "in a way not seen since 1945."

The Bigger Picture: Government Investment in AI Infrastructure

UKPoliticsDecoded was in the process of writing an article on the govevernments AI initiatives when the Government announced this procurement reform.

The national security procurement announcement is directly linked to a much larger government initiative, a sustained programme of public investment in artificial intelligence infrastructure. The scale of that commitment, set out across several government publications and press releases over the past year, is significant. Below is what the Government has committed and what it means in practice.

AI Growth Zones: £28.2 Billion in Investment, 15,000 Jobs

The Government has designated five AI Growth Zones across Great Britain, with two in Wales and one in Scotland. These are specific locations where the Government is streamlining planning rules and improving energy access to make it faster and cheaper to build large AI data centres.

According to the Government's own one year update on its AI Opportunities Action Plan, these five zones have together generated £28.2 billion in private investment and are creating more than 15,000 jobs. Each zone also receives up to £5 million of targeted public funding to drive local AI adoption and skills.

The Government claims the overall package of reforms for AI Growth Zones including faster grid connections and energy pricing support could reduce the time it takes to connect a new data centre to the power grid by up to five years, and save a 500 megawatt data centre up to £80 million per year in electricity costs. The Government has projected that the AI Growth Zone programme could unlock up to £100 billion in additional investment overall.

The North East: Up to £30 Billion and 5,000 Jobs

The North East of England has been designated as an AI Growth Zone. The Government announced in September 2025 that this single zone could attract up to £30 billion in investment and create more than 5,000 jobs. Of the £30 billion, £10 billion had already been committed by investment firm Blackstone for a hyperscale data centre campus in Northumberland.

The North East zone also became the site for Stargate UK a partnership between British firm Nscale, American companies OpenAI and NVIDIA, to build new AI computing infrastructure. The first phase involves OpenAI taking up to 8,000 GPUs (graphics processing units, the specialist chips that power AI systems), with the potential to expand to approximately 31,000 GPUs over time. Nscale is also building what it describes as the most powerful supercomputer in the UK, in partnership with Microsoft.

The AI Growth Zone in the North East is expected to increase total energy capacity to 1.1 gigawatts over the next six years, making it one of the largest data centre sites in Europe.

Electricity Subsidies for AI Data Centres

One of the most concrete financial commitments in the AI Growth Zone programme is a targeted electricity pricing discount for data centres located in specific areas of the UK. From April 2027, subject to legislation, data centres in eligible AI Growth Zones will receive reductions in their electricity costs of:

The rationale for these discounts is that data centres located in Scotland and the North of England can absorb surplus electricity generated by wind farms and other renewable sources in those regions, electricity that would otherwise be wasted or constrained on the grid. By locating in those areas and using that excess power, the data centres would reduce overall system costs. The Government says the discounts reflect those savings, and will not add costs for other electricity bill payers.

£2 Billion to Expand Public Computing Power Twentyfold

Alongside private investment, the Government has committed £2 billion of public money to expand the UK's own computing capacity the processing power needed to train and run advanced AI systems. The stated aim is to increase public sector compute capacity twentyfold by 2030.

As part of this, the Government launched the Isambard AI supercomputer at Bristol University in July 2025, and earmarked up to £250 million to scale up cloud computing capacity for the AI Research Resource, a facility that provides free access to computing power for British researchers, businesses, and start-ups.

In January 2026, the Government also confirmed it would fund a sixfold increase in supercomputer capacity at the University of Cambridge by Spring 2026. And £750 million has been committed to fund the next national supercomputer, which will be located in Scotland and linked to the Edinburgh Parallel Computing Centre.

£500 Million Sovereign AI Unit

The Government has established a Sovereign AI Unit, backed by up to £500 million of public funding. The unit's purpose is to invest in and support British AI companies, particularly those working in areas the Government identifies as critical parts of the AI value chain. The next phase of the unit is due to launch in April 2026, and will be chaired by James Wise of Balderton Capital.

So far, the unit has allocated computing resources to support AI researchers and high growth start-ups, and invested £8 million in seed funding for a consortium working on drug discovery datasets.

£100 Million National Data Library

The Government backed a National Data Library with over £100 million at the 2025 Spending Review. The library's purpose is to make government and public sector data more accessible for AI development creating what the Government describes as high value data assets that can help researchers, businesses, and public services apply AI more effectively.

£600 Million Health Data Research Service

The Government and the Wellcome Trust have committed up to £600 million to create a Health Data Research Service, a single secure access point to regional and national health datasets. The stated aim is to accelerate medical research and improve diagnosis by allowing AI systems to be trained on large scale health data in a controlled and secure environment.

£187 Million Skills Programme

The Government has committed £187 million to a programme called TechFirst, which aims to open pathways for young people into the technology sector. This includes TechGrad undergraduate scholarships for students in AI and related fields, and the Spärck AI scholarship to attract exceptional postgraduate students to leading UK universities. The wider ambition is to provide 10 million workers with AI skills by 2030, with over one million AI training courses already delivered.

£240 Million for AI Safety Research

The Government has backed the AI Security Institute (AISI) with £240 million at Spending Review 2025. AISI now has over 100 researchers and has tested 30 frontier AI models. Its role is to evaluate the risks and capabilities of the most advanced AI systems and to shape international standards for how those systems are assessed and controlled.

UK-EU Collaboration on AI: The AI Factory Antenna

Alongside domestic investment, the Government has been expanding international collaboration on AI computing. The UK joined the EuroHPC Joint Undertaking in May 2024, which gives UK researchers free access to some of Europe's most powerful supercomputers through the UK's association with the Horizon Europe research programme.

In May 2025, building on the UK's new trade deal with the EU, the Government launched a process to select a UK organisation to host an AI Factory Antenna, a facility that would act as the UK's gateway to dedicated AI supercomputing sites across Europe known as AI Factories. The Government allocated up to £2.5 million of its own funding for the application, which would be matched by up to €5 million from EuroHPC if the application succeeded.

Minister for AI Feryal Clark said:

"Supercomputers are the turbo chargers of discovery. By strengthening our partnership with Europe, we're giving British innovators the compute power to solve climate and health challenges, grow the economy, and deliver our Plan for Change."

Conclusion

Taken together, these announcements amount to a major shift in how the UK Government is using public money to shape the country's AI and industrial future. The scale of the figures involved is large, and it is worth being clear about what they represent.

Some of the money, such as the £28.2 billion linked to AI Growth Zones is private investment attracted by government policy, not direct public spending. The Government created the conditions (easier planning, electricity discounts, designated zones) and private companies have committed capital as a result.

Other figures are direct public spending, the £2 billion for computing capacity, the £500 million Sovereign AI Unit, the £187 million skills programme, the £240 million for the AI Security Institute, and the £100 million National Data Library are all funded by the taxpayer.

The electricity subsidies are a third category, they are discounts on future bills, not grants, and their cost is designed to be offset by the savings the data centres create in the wider electricity system by absorbing surplus renewable power.

The procurement changes announced on 26 March 2026 sit alongside all of this. By requiring departments to buy British in key sectors, and to think twice before outsourcing, the Government is attempting to use the power of public purchasing worth hundreds of billions of pounds across the public sector each year to reinforce its industrial and AI ambitions.

Almost as an afterthought, with this amount of public funds been invested into AI, could it have been better spent? Maybe if a fraction of these funds realocated into publically owned offshore windfarms to create competition in the sector and bring down household and business electricity bills?