Built by graduates: The UK universities creating the next generation of AI businesses

Artificial intelligence is no longer the preserve of big corporations. Small businesses are increasingly using AI to automate admin, better understand their customers and develop new products, helping to close the innovation gap with larger businesses that have traditionally had greater resources to invest in new technology. 

At the same time, AI is lowering some of the barriers to starting a business, allowing founders to build, test and market ideas with fewer people and potentially less capital than would previously have been required. 

Small businesses can use AI to boost business growth by leveraging it for a variety of benefits. Matt Ovenden launched The Wine Engine, an AI-driven platform that tailors wine recommendations to individual tastes, while maintaining that joy of discovery. Ask Silver is a scam checker that’s free-to-use and uses AI to highlight potential scams. 

The investment market is already reflecting this shift. In 2025, AI businesses accounted for a record 44% of UK smaller-business equity investment, according to the British Business Bank. 

The UK government is also putting greater emphasis on AI adoption among SMEs, with an ambition for UK small and medium-sized businesses to become the most digitally capable and AI-confident in the G7 by 2035. 

But for the UK to turn AI investment and adoption into a new generation of businesses, it needs more than investment capital and technology – it needs people who know how to build with it. 

Universities and further education institutions are therefore a vital part of the UK’s AI economy. Much of the foundational research that commercial AI builds on takes place in universities, where researchers can pursue long-term, high-risk projects that companies may be less willing to fund.

Universities can also host or help to run national research assets, including supercomputing facilities such as Isambard-AI at Bristol, the UK’s fastest and most powerful supercomputer, which has been purpose-built for AI research.

They also help with policy-making. Universities often work and collaborate with businesses on applied projects, and help sectors like the NHS, finance and manufacturing to adopt AI. 

In short, universities are supplying the people, ideas and infrastructure the sector relies on, creating the building blocks for future innovation and economic growth. 

Which universities are building the UK’s AI business pipeline? 

To find out where the next-generation of the AI workforce is strongest, money.co.uk analysed 146 UK universities across three areas: AI-related education, graduate employment and university spinout activity. 

Rather than looking simply at just how many AI courses a university offers, the index considers whether universities are developing AI and technical skills and whether that expertise is translating into employment and new business. 

The analysis looked at: 

• AI course volume: the number of courses containing AI in their title, alongside related disciplines including computer science, robotics, machine learning, automation and data science.

• Graduate employability: the proportion of graduates from computing, engineering, technology and mathematical science subjects entering full-time employment.

• Spinout activity: the number of active university spinouts in engineering, physical sciences and mathematics, alongside businesses identified as specifically AI-related through their company name or .ai domain. 

Together, these measures provide a picture of where universities are not just teaching AI, but helping to create the talent and businesses that could underpin the UK’s next phase of growth. 

The top 20 most AI-ready universities

• 1. Imperial College London — 31 AI courses (10 specific) | 80 STEM spinouts (3 .ai) | 71% FT employment

• 2. University of Leeds — 31 AI courses (10 specific) | 34 STEM spinouts (1 .ai) | 70% FT employment

• 3. University of Bristol — 27 AI courses (9 specific) | 51 STEM spinouts (2 .ai) | 64% FT employment

• 4. University of Birmingham — 27 AI courses (10 specific) | 26 STEM spinouts (1 .ai) | 66% FT employment

• 5. University of Liverpool — 26 AI courses (12 specific) | 19 STEM spinouts (1 .ai) | 61% FT employment

• 6. Loughborough University — 20 AI courses (8 specific) | 13 STEM spinouts (1 .ai) | 65% FT employment

• 7. University of Edinburgh — 26 AI courses (7 specific) | 31 STEM spinouts (1 .ai) | 63% FT employment

• 8. University of Surrey — 20 AI courses (11 specific) | 23 STEM spinouts (1 .ai)| 60% FT employment

• 9. University of Nottingham — 14 AI courses (6 specific) | 24 STEM spinouts (2 .ai) | 64% FT employment

• 10. University of Oxford — 15 AI courses (4 specific) | 123 STEM spinouts (6 ..ai) | 65% FT employment

• 11. University of Dundee — 19 AI courses (12 specific) | 8 STEM spinouts (1 .ai) | 58% FT employment

• 12. University of Warwick — 15 AI courses (4 specific) | 34 STEM spinouts (2 .ai) | 66% FT employment

• 13. University of Cambridge — 12 AI courses (5 specific) | 174 STEM spinouts (10 .ai) | 62% FT employment

• 14. Queen Mary, University of London — 24 AI courses (12 specific) | 17 STEM spinouts (1 .ai)| 50% FT employment

• 15. University of Strathclyde — 21 AI courses (3 specific) | 14 STEM spinouts (1 .ai) | 65% FT employment

• 16. Newcastle University — 16 AI courses (4 specific) | 20 STEM spinouts (1 .ai) | 63% FT employment

• 17. University College London (UCL) — 36 AI courses (9 specific) | 52 STEM spinouts (0 .ai) | 70% FT employment

• 18. Heriot-Watt University — 13 AI courses (2 specific) | 17 STEM spinouts (2 .ai) | 68% FT employment

• 19. City, University of London — 20 AI courses (7 specific) | 9 STEM spinouts (1 .ai) | 47% FT employment

• 20. Cardiff University — 10 AI courses (3 specific) | 11 STEM spinouts (1 .ai) | 70% FT employment

Imperial leads on breadth and business potential 

Imperial College in London takes the top spot, with a strong performance across all three measures rather than relying on a single area. 

It has 31 AI-related courses, including ten specifically focused on AI, alongside 80 science, engineering and mathematics spinouts. Its graduates also record the highest full-time employment rate among the universities in the top 10, at 71%. 

This combination is significant because it illustrates the different stages of the AI business pipeline: education, talent and commercialisation. 

A university can produce large numbers of technically skilled graduates without necessarily producing new businesses. Equally, it can have a strong spinout ecosystem without having the same breadth of AI education. 

The universities appearing at the top of the index demonstrate strength across multiple stages of business development for the next generation. 

The biggest AI course provider isn't necessarily the strongest commercial ecosystem

UCL (University College London) has the highest number of AI and AI-related courses at 36, yet sits at 17 in the overall table, due to other universities faring better in other parts of the analysis. 

By comparison, Imperial has fewer AI and AI-related courses (31), but has 80 relevant spinouts, including three identified as AI-specific, compared with UCL’s 52 relevant spinouts and no AI-specific spinouts. This suggests that having a larger number of AI-related courses doesn’t necessarily translate directly into greater AI commercialisation.

Oxford and Cambridge have huge spinout ecosystems, but their AI course provision isn't the largest

Oxford has 123 relevant spinouts and Cambridge has 174, dwarfing most other universities in the dataset. However, Oxford only has 15 AI/AI-related courses, while Cambridge has just 12. 

Some universities are taking a more specialist approach to AI education

The University of Liverpool has 12 specifically AI-focused courses out of 26 AI and AI-related courses, while Dundee has 12 out of 19, and Queen Mary has 12 out of 24. This places all three among the universities with the strongest concentration of specifically AI-focused courses in the dataset. 

The most AI-ready universities in London

• 1. Imperial College London — 31 AI courses (10 specific) | 80 STEM spinouts (3 .ai)| 71% FT employment

• 2. Queen Mary University of London — 24 AI courses (12 specific) | 17 STEM spinouts (1 .ai) | 50% FT employment

• 3. University College London — 36 AI courses (9 specific) | 52 STEM spinouts (0 .ai) | 70% FT employment

• 4. City, University of London — 20 AI courses (7 specific) | 9 STEM spinouts (1 .ai) | 47% FT employment

• 5. King’s College London — 11 AI courses (5 specific) | 11 STEM spinouts (1 .ai) | 58% FT employment

• 6. Royal Holloway, University of London — 17 AI courses (4 specific) | 1 STEM spinout (0 .ai) | 62% FT employment

• 7. Birkbeck, University of London — 17 AI courses (7 specific) | 0 STEM spinouts (0 .ai) | 50% FT employment

• 8. Goldsmiths, University of London — 11 AI courses (4 specific) | 2 STEM spinouts (0 .ai) | 56% FT employment

• 9. Brunel University of London — 16 AI courses (5 specific) | 0 STEM spinouts (0 .ai) | 49% FT employment

• 10. University of Greenwich — 15 AI courses (4 specific) | 1 STEM spinout (0 .ai) | 47% FT employment

While only four London universities appeared in the national top 20, the capital has a strong ecosystem for developing AI talent and innovation. 

Queen Mary University of London ranks second, supported by 12 specialist programmes and 17 STEM spinouts. Overall, the figures indicate that London’s strength extends beyond education alone, with graduate outcomes and links between universities and commercial innovation also contributing to institutional performance.

AI investment could create opportunities far beyond the technology sector 

The implications extend beyond companies whose core product is AI-based. AI could also lower the cost and complexity of starting a business across sectors, from automating time-intensive tasks to reducing the need for expensive software and external services. 

This could make business loans and other forms of business finance go further, giving entrepreneurs more scope to invest borrowed capital into growth, equipment, marketing and hiring. 

As a result, we could see more small businesses using AI as infrastructure. An AI startup, for example, might build a new model, software product or AI application, whereas an AI-enabled small business might be a traditional business that uses AI to operate more efficiently, launch faster or compete with much larger companies. 

The latter of which could ultimately be just as important to the UK’s productivity and business landscape. ONS data shows that AI adoption among UK businesses with 10 or more employees increased from around 12% in late 2023 to around 35% by 2026, although adoption remains relatively shallow, and larger businesses are more likely to have adopted the technology. 

That leaves a significant opportunity for smaller businesses, particularly if they can access the skills and expertise needed to implement AI effectively.

The UK’s next generation of founders may be learning about AI right now

The UK’s AI opportunity is not simply about attracting investment into existing AI companies. It’s also about creating the conditions for thousands of future founders and small business owners to understand what AI can do for their businesses. 

A computer science graduate could build an AI startup, but they could equally become the founder of a new software consultancy, launch an AI-enabled recruitment business or create a technology product for an established industry. 

But a graduate doesn’t necessarily have to build an AI company at all. Someone with experience in AI, data or automation could enter an existing sector and identify an opportunity to make an inefficient process dramatically cheaper, faster and more effective.

In this sense, universities are liable to become an important part of the UK’s AI entrepreneurship ecosystem, supplying not only employees for established businesses, but potential founders, technical co-founders and the skills needed to build AI into businesses throughout the economy. 

Where does this leave the UK’s small business economy? 

The UK's AI opportunity could ultimately be determined by what happens outside the country's biggest technology companies.

The government has identified SME adoption as an important part of the UK's AI growth strategy, while its 2026 update on AI adoption highlights barriers including capability, cost and awareness among smaller businesses. Universities can potentially address the first of those barriers by creating a pipeline of people who understand the technology. That could benefit existing businesses, but it could also help create new ones.

While universities can help to address growing skills needs, businesses may also need access to finance to invest in new technology, training and infrastructure, through small business loans. 

As AI tools become increasingly accessible, the advantage may shift from simply having access to AI to knowing how to apply it to a commercial problem.

The universities producing graduates with those skills, while also supporting spinouts and commercialisation, could therefore play an important role in determining what the UK's next generation of small businesses looks like.

Methodology 

money.co.uk looked at the 146 UK universities, based on institutions from the official Universities UK recognised bodies register, alongside other institutions that legally hold university status. 

We analysed AI and related courses, using UCAS data, to identify all courses held at each university that reference AI/Artificial Intelligence, machine learning, data science/scientists, robotics and automation in the course name. 

Where the same course appeared multiple times because of sandwich years, industrial placements or part-time study options, these were removed to avoid duplication. Where courses were offered at multiple levels or were otherwise distinct qualifications, they were retained as separate courses. 

HESA (Higher Education Statistics Agency) data was used to assess graduate outcomes, using the most recent available dataset. We looked at graduates in computing, engineering and technology, and mathematical sciences, as these subject areas encompass the majority of AI-related study. The percentage in full-time employment was calculated based on the total number of respondents who provided an outcome. 

We also analysed HESA’s University Spin-out register to identify spinouts that are currently active and classified under engineering, physical sciences or mathematics, providing the total number of relevant spinouts associated with each university.

Alongside this, we identified AI-specific businesses associated with each university, defined as businesses with a .ai domain or with “AI” or “artificial intelligence” in their name. 

Each data point was assigned a weighting out of 25, with the exception of general spinouts, which were weighted out of 10, to create an overall index score out of 110.