Talk to Sales
Client Log In →
UK Public Sector AI Procurement Tracker

UK Public Sector AI Procurement Tracker

Posted by James Piggott Picture of James Piggott on 04 November 2025
Report: UK Gov's 39 'Strategic Suppliers', analysed
New Nav Bar, Insights - Strategic Suppliers 2025 v2
Download report

Contents:

Last updated: 4th November 2025

Sky News AI Tracker Thumbnail v2

NB: On November 4th, Sky News' Sam Coates used data from this Tracker to report on the government's £3.3bn worth of awarded AI contracts. Watch his full explainer here.

Introduction


Artificial Intelligence (AI) is capturing the world’s imagination, and that includes the UK public sector.

Public bodies have been purchasing AI-based solutions for well over a decade, but recent breakthroughs - specifically in Generative AI since the launch of ChatGPT in late 2022 - have spurred a fresh wave of interest in how government could benefit from this rapidly-evolving technology.

Whilst loud voices debate the future of AI, we at Tussell wanted to cut through the noise and look at the facts.

We’ve set out to answer a simple question: How much has the government spent on procuring AI so far, and what is it buying?

Introducing: Tussell's AI Procurement Tracker.

Inspired by an earlier tracker we produced to measure Covid-related procurement contracts, this new Tracker draws on data from our own market intelligence platform to map the landscape of AI adoption across government.

Our AI Procurement Tracker is your go-to source of fact-checked data to understand which AI capabilities the government is buying, who from and at what cost.

This will be essential reading if:

  1. Your company supplies tech to government and you want to follow the latest market trends, or

  2. You work in a public body and want to know how your peers are
    adapting to rapid change, or

  3. You are a journalist, commentator or policy advisor in search of

    hard facts to back up your research.

 

The AI Procurement Tracker


Between 01/01/2018 and the 20/10/2025:

The UK public sector awarded

1,565 AI contracts


In total, they were worth

£3.35 billion

 

New call-to-action

 

November 2025 Update: key takeaways

As of late October 2025, this year has officially become the biggest on record for AI-related public sector contracts, with 366 awards - a number that’s likely to climb further before year-end.

Excluding Microsoft’s £1 billion contract in 2021, 2025 also ranks as the top year by total contract value for AI-related deals.

Sky News' Sam Coates reported on our latest findings in an exclusive piece, outlining where the government is - and isn't - procuring AI. 

Some notable contracts awarded since our last update include:

  • In September, IBM was awarded a £9 million 'Nexus AI' contract by the DWP, "related to the exploration, deployment, and ongoing support of Artificial Intelligence (AI) systems and services".

  • CDW secured a £6.75 million contract to provide ChatGPT licenses to the MoJ.

 

*

 

Chart 1: Value and volume of awarded contracts, since 2018


This chart shows that both the number and value of AI contracts have grown significantly since 2018, with more contracts awarded so far in 2025 versus any prior year.

Barring a spike in 2021 - caused by a £1 billion 'Supercomputing 2020+' contract, awarded by the Met Office to Microsoft - the total value of AI contracts has shown steady, strong growth between across the period, growing 1,113%.

 

👉  To download a sample of the underlying data, click here.

 

Chart 2: Largest buyers of AI by contract value, since 2018 

Top AI Buyers - Nov 2025 v1


Chart 3: Largest suppliers of AI by contract value, since 2018

Top AI Suppliers - Nov 2025 v1

 

Chart 4: Breakdown of AI-related contract value by theme

The UK public sector is harnessing AI to pursue a wide range of objectives - and this diversity is reflected in the variety of contracts being awarded.

To better understand where government procurement is focused, we asked ChatGPT to categorise these contracts using taxonomies derived from the three strategic pillars of the AI Opportunities Action Plan.

Please note this AI categorisation is still a work-in-progress, and results may change as we improve the model.

Here are the results:

 

Below, we've spotlighted some key contracts from these categories to illustrate what each means.

  • Compute & Infrastructure - focuses on providing the physical and digital backbone for AI - from high-performance computing and cloud platforms to sovereign infrastructure that ensures secure, scalable access to compute resources (e.g. the Met Office's £1bn Supercomputing contract)

  • Adoption & Deployment - contracts which promote the piloting and scaling of AI projects, and encouraging uptake across society (e.g. National Highway's £35mn contract with Deloitte to introduce a 'Digital Lab' in their Chief Data Office)

  • Innovation & Frontier AI - supports cutting-edge AI research and the development (e.g. IBM's £13.5mn contract to deliver 'GenAI Lighthouse Projects' within the DWP)

  • Data Management & Enablement - covers unlocking, standardising and sharing high-value datasets so that AI models can be developed and deployed responsibly (e.g. NHS England's £182mn contract with Palantir to provide a 'Federated Data Platform')

  • Governance, Management & Regulation - contracts that promote the safe and ethical use of AI in society (e.g. Anmut Consulting's £25mn contract with National Highways to develop their Chief Data Office's 'Data Governance & Assurance' division, including their AI Governance Strategy)

  • Skills & Workforce - up-skilling staff to ensure they can properly harness the potential of AI in their work (e.g. Kainos' £7.4mn contract with the MOD to provide staff "the capacity and skills required to deliver AI and AS concepts sustainably and at scale")

  • Market Confidence & Public Trust - contracts that promote transparency, ethics and public engagement around AI.

 

 

Chart 5: Top 5 AI frameworks (by value of call-off contracts)

AI Top Frameworks Nov 2025 v1

Many of the identified AI contracts were awarded via a framework agreement or Dynamic Purchasing System.

These preferred-supplier agreements enable contracting authorities to expedite their procurement, by publishing and awarding tenders to a closed group of pre-vetted suppliers.

Tussell's framework intelligence shows that just over a third of the contracts awarded since Jan 2018 were called-off via a framework agreement.

The wider public sector, by comparison, awards roughly 1 in every 4 contracts via a framework, meaning buyers seem to be harnessing frameworks more frequently when purchasing AI-related services.

 

👉  Want to dig into the data for yourself? Download a sample here.

 

New call-to-action

What is the data telling us?

We think that the amount of money spent so far on AI is big, but not as big as the hype surrounding Gen AI would suggest.

£3.35 billion sounds like a lot, but remember: just under a third of this stems from a single £1 billion contract.

But even when taken at face value, £3.35 billion still only accounts for roughly 2.5% of the total value of all IT Services and Software contracts awarded by the public sector over the Jan 2018 - Dec 2024 period (£134bn)

The rate of increase in the volume and value of AI-related contracts each year will likely mean this proportion will grow over the coming decade - but, for the time being, AI-related procurement remains a small sub-section of government tech spending.

 

What do the experts think?


We approached two domain experts for their comments on the data – Joe Hill, Policy Director at leading think tank ReState, and Sean Williams, Founder and CEO of AutogenAI.

Asked to reflect on the AI Procurement Tracker's findings, Joe Hill pointed out that many contracts have been awarded to non-specialist firms, large digital companies and systems integrators:

"This shows the government has begun its AI adoption journey, and that it's largely buying technical support rather than buying software directly."

As public sector familiarity with and trust in AI increases, experts are predicting a continued uptick in AI-related procurement. Sean Williams says:

"Artificial intelligence spend is only going to increase over the next months and years.

"Generative AI is a paradigm shifting general purpose technology - like the printing press, electricity or the internet. We are in the equivalent of 1995 for the internet with generative AI. Over the next decade it will transform entire tranches of government, business and society.

This report shows the first stirrings of the change that is coming.”

Reform’s Joe Hill has a similar prediction. He says we'll "see a growth in AI-related procurement, and probably a shift from procuring external support to building bespoke solutions, to buying off-the-shelf software and using open-source products."

Our AI Procurement Tracker will be updated monthly to monitor trends and to highlight interesting themes.

Each month we will also invite experts to interpret the data and use it to try to look around the next corner. If you would like to contribute to the discussion, get in touch!

 

New call-to-action

 

What's the methodology?

 

The AI Tracker draws on open public procurement data, aggregated by Tussell’s market intelligence platform.

We have filtered all awarded contracts in the period January 2018 to October 2025 for only those containing keywords in their title or description that are capabilities, products or companies.

For the categorisation of the AI contracts we've used ChatGPT with human oversight.

 

AI Tracker - in the media

 

We're proud that the national media has utilised Tussell's AI Procurement Tracker to report on the state of AI procurement in the UK public sector. Examples include:

 

*

 

Measuring what is and isn’t AI is highly subjective.

We stand by our prudently narrow definition based on explicitly AI-related keywords, but we will continue to iterate on our exact definition as time goes on.

If you have any questions or comments about how the AI Procurement Tracker was compiled, email us at contact@tussell.com

 

See how we uncovered these insights