The UK government has launched a £100 million competition to help British AI companies develop and test technologies for major public-sector challenges. The first projects target NHS productivity, computing efficiency, defence systems, and AI security, with the government aiming to improve services while helping domestic firms scale internationally.
Key takeaways
The programme is designed to give promising companies an early public-sector customer and a route beyond prototype development.
- Four initial challenges cover healthcare, AI computing, defence, and cyber resilience.
- Upfront payments may help smaller firms compete with larger suppliers.
- Companies will retain the intellectual property they develop.
- The government plans further competitions under the Sovereign AI scheme.
Four challenges for British AI companies
The first competitions focus on technologies that are developed enough for real-world demonstrations but still need funding and operational access to prove their value.
The NHS productivity challenge will seek systems that automate workflows, coordinate care and support clinical decision-making. The objective is to improve productivity and staff experience while contributing to the ambitions of the NHS 10 Year Health Plan.
A second challenge, led by the Department for Business, Innovation, Science and Technology and ARIA’s Scaling Inference Lab, will target more efficient AI computing. The work is intended to lower costs and support the expansion of the UK’s public AI compute capacity for researchers and businesses.
The Ministry of Defence will invite solutions that securely connect data and advanced AI capabilities across defence environments. A separate competition, developed with the National Cyber Security Centre, will focus on testing and strengthening the security of increasingly capable AI agents.
A different route into government procurement
The scheme is intended to address a familiar obstacle for startups: public contracts often favour companies with substantial turnover, cash reserves and established delivery records. The government says upfront funding may be available where appropriate, allowing smaller British firms to build and demonstrate their systems without carrying the full financial burden.
Successful companies will retain the intellectual property created through their projects. That could enable them to turn a government pilot into a commercial product for wider use in the UK and overseas—a key issue for the domestic AI market and one that Power AI Media will continue to track across its technology and policy coverage.
Sovereign AI expands its role
The competitions are the first phase of the Sovereign AI R&D Procurement Scheme. The initiative is intended to operate with the speed and risk appetite of an investment fund, while using the state’s purchasing power to help strategically important companies start, grow and compete from the UK.
Participating firms will work with government departments on demonstrator-stage technologies. If successful, those projects could be adopted more widely across the public sector and help create reference customers for global expansion.
Wider international cooperation planned
The announcement also included plans to open the AI Economics Institute to cooperation with G7 countries. The institute is intended to improve understanding of AI’s economic effects and provide policymakers with evidence about its opportunities and risks.
For the UK, the combined approach is both industrial and public-service focused: use government demand to support domestic AI companies, while building stronger evidence and safeguards around adoption. The first competitions will test whether that model can deliver practical improvements beyond the hype surrounding the technology.