In a special quarterly edition of GovTech in Focus, PUBLIC – in partnership with Tussell, the data provider on UK government contracts and spend – looks at public-sector spend on particular technologies, including value, trends and key contracts. In this edition, we investigate how the public-sector is buying and using artificial intelligence, and what this means for startups. 

Ten years ago, entrepreneurs pitching AI innovations for use in public services would have received little positivity from buyers. AI was – in the mind of many government digital teams – untested, unproven and therefore unappealing as a tool for governing. 

Flashforward to 2020 and the use of AI in government and public services is now common parlance, as those in power have caught on to the potential of these technologies to introduce major efficiencies into a number of core government processes. With AI, information that might once have taken thousands of hours to process – and develop actions based on the resulting insight – now takes just hours, even minutes. As these technologies continue to develop, so too does their value as tools for speeding up policy-making, diagnostics and analysis, process flows, and other procedures that the public sector undertakes daily. 

As the public sector moves more of its services online, AI technologies are forming the basis channels for citizens to engage and communicate with government.  And for new innovations in the fields of robotics and autonomous vehicles – AI is the backbone of a thousand other tools that are likely to change the world we live in. 


What is Artificial intelligence and how can it be used by Government and the public sector?

Simply put, artificial intelligence isn’t just one technology; it’s an umbrella term for dozens – if not hundreds – of tools that vary in purpose and scope depending on the degree of sophistication in the way they interpret and learn from data input. 

To research this article, PUBLIC and Tussell went digging through an extensive backlog of contract data to find out where these tools have made their mark on the public sector. From ‘Artificial Intelligence’ to ‘Turing test’,  ‘algorithm’ to ‘predictive analytics’, we were determined to look under every rock.

In total, we searched for 61 different technologies to find out what technologies are being used in Government and what they’re being used for – whether it’s data analytics, research, delivery or even policy-development!


Getting to know the startups: Faculty – Artificial Intelligence made real

Formerly known as ASI Data Services, Faculty relaunched early last year with a powerful rebrand that emphasised its extensive efforts in the proliferation of AI technologies across multiple industries.

The company is a leader in the delivery of AI software, strategy and skills to both business and public sector players. Notable projects it has run in the GovTech space include a collaboration with the UK Home Office Counter-terrorism Unit, which involved using the company’s deep learning and Natural language processing technologies to stop the spread of Daesh propaganda. 

More recently, the company has been working with NHSX (the innovation arm of the UK’s National Health Service), along with big tech companies such as Palantir and Microsoft, using data to assist the NHS in understanding more about the spread of COVID19 in the UK. 

How is the Government helping to encourage the use of AI in public services?

AI has taken a central position in the UK Government’s approach to new technologies. The UK’s industrial strategy highlighted both the potential of the technology to increase productivity and as a cornerstone for the UK’s digital economy – outlining its intent to put the UK at the forefront of the artificial intelligence and data revolution.

This aim culminated in the launch of the AI Sector Deal and a new Office for AI responsible for its implementation. That body, along with the AI Council and the Centre for Data Ethics and Innovation (CDEI) have worked hard to further promote AI’s value as well as to agree on standards for both the use of the data being analysed as well as standards for public bodies procuring relevant technology. or using AI to assist decision-making. The work of these agencies so far has been impressive, and has established the UK as one of the world’s leading governments for policy and strategy on AI.


But what does AI procurement look like in reality – is the public-sector really buying AI?

By 2020, the use of AI by the government and the public-sector has seen a huge jump.  

Chart: Number and value of public-sector AI contracts since 2015

Thanks to amazing developments in the field, and a general shift in the readiness and technical maturity of public sector buyers, the day-to-day design and delivery of a number of services and processes is quietly driven by the outputs of AI tools. In the five years since 2015, there have been 111 contracts awarded to suppliers of different technologies, amounting to a total of £70m public-sector spend on artificial intelligence. 


What problems is AI solving in the public-sector – and for who?

But this level of innovation is new for many public-services and still isn’t a universal approach, as the graph below shows. 

Graph: Number of Public Sector AI Contracts signed since 2015 by Central Government, Local Government, NHS and other public bodies


Until 2018, almost all spend in AI originated at central government level. Even now, Whitehall continues to be the leading awarding body. 81 of the 111 contracts awarded came from government departments, with the Ministry of Defence awarding the most, followed by UK Shared Business Services – a jointly-owned subsidiary of the Department for Business Energy Industrial Strategy (BEIS) and UK Research and Innovation (UKRI) that specialises in procuring HR & Payroll, Finance, Procurement and IT services on behalf of UK research agencies.

A scan through the data finds that the majority of the contracts awarded pre-2018 relate to data analytics; keywords such as ‘big data’ and ‘data science occurring regularly, such as these two high value contracts for the Home Office and HMRC. 

Notable exceptions prior to 2018 include this fascinating UK Shared Business Services  contract worth £13.5m:  issued on behalf of The Hartree Centre of the Science & Technologies Facilities Council, UKSBS sought to procure a compute cluster and its associated software and services for the Snowdonia Project as it moved to a more balanced portfolio of HPC, HPDA, Big Data and Machine Learning tools and research.

Since 2018, there’s been a gradual expansion in the types of problems that central government has been seeking to solve through AI; the technology is no longer just about sorting through the data, it’s about selection, production, and even decision-making. 

Beyond central government, other public bodies have been a gradually increase source of key contracts; although the increase here has been slower, it’s relatively widespread – with high-value contracts emerging across local government, healthcare providers and in universities and other innovation hubs. 

So who’s using AI, and what does this tell us that can be useful for developers considering working with the public-sector?

  1. AI in Local Government:

Like others outside of central government, Local Government spend on AI is relatively new. In a short time though, the tenders we’ve seen have been for a diverse range of technologies. Here are our favourites:

Note on the above: we checked – the chatbot ‘’ is still live, so do visit Stockport’s website to see what it can do.


  1. NHS: what are they using AI for?


  1. Academia & Research: what are they using AI for?

Again, a slow growing space for AI – perhaps surprisingly – but academic and research has generated some of the most fascinating potential uses for AI over the past 5 years, including:


So who’s winning AI contracts and what does this mean for SMEs?

In the past 5 years, PUBLIC has seen (and sometimes backed!) dozens of fantastic startups using AI and related technologies to transform public services. From chatbots, to diagnosis, patient management and beyond, there are a massive range of possible applications for artificial intelligence technologies in improving citizens’ lives. 

Getting to know the startups: FUTR

FUTR (  is a machine learning platform which allows chat bots to augment internal and external systems and deploy two-way natural language conversation.

In practice, this means deploying intelligent chatbots across 1000s of organisations in both the public and private sectors – and allow users to communicate with colleagues instantaneously in over 170 languages. 

FUTR has made waves in the industry and is an alumni of the 2018 GovStart Programme. Their Chatbots can be found across the web as tools for both employee engagement and consumer communication. Most recently, they provided the Chatbot interface that assisted users of – the website for PUBLIC-NHSX COVID-19 innovation challenge – and have offered their services to local authorities throughout the crisis. 

Good news then for these startups, then – our data shows that SMEs have won just over half of all the AI contracts awarded by the public sector since 2015. 

Proportion of public-sector AI contracts won by SMEs since 2015

Our experience in public sector tech procurement has tended to show that finding a place for truly innovative technology in government is difficult, largely due to a preference among buyers for known quantities – including tested technologies and more established outsourcing providers. It’s fascinating to see then, that small businesses – such as startups – are major providers of AI to the public sector.

Those thinking that SMEs may win more contracts, but that the likes of IBM and Atos will surely have the most valuable contracts are in for a surprise too, with SMEs making up half of the top 20 AI suppliers to the public sector by value. 

Graph: Top Suppliers of AI to the public-sector since 2015

While Atos – a major supplier of IT services – takes the top spot in the top 10, based on its involvement with the £13.5m Hartree contract, contracts provided by SMEs such as Hippo Digital, 6 Point 6, FlatworldWorks and Fordway Solutions account for over £15m of public-sector spend on AI tech. These include the Home Office Strategy Contract mentioned earlier in the article, as well as other high value contracts including for the Department for Education.

However, a lot of the contracts at this level focus on recruitment, and it’s not until we get to the 11-20 spots that we see startups and SMEs providing real innovation in the public sector, with contracts worth between £250k and £1m awarded to startups by MoD, MHCLG, Oxfordshire County Council and several others.


Conclusions: “I’m an AI startup – what do I need to know about selling to the public sector?

For startups, there’s a lot to digest when considering how to sell AI to government. If one thing seems to be true that is  often not the case elsewhere, AI companies aren’t necessarily solving one solution – they have a product and it can be adapted or developed to suit other needs – so consider how you can pivot to take advantage of new tenders for AI technology as they arise.

The good news is that the environment for AI startups is improving – the various bodies tasked with looking into how to use AI in government aren’t stifling innovation; they’re providing guidance that will help uncertain public buyers become more comfortable with using AI day-to-day. However, as a supplier, it’s useful to familiarise yourself with the guidance and frameworks so that you can be on the same page as those to whom you’re trying to sell. 

Speaking of frameworks – a recent announcement makes us think it’ll soon be even easier for AI companies to engage with buyers and vice versa: A new Dynamic Purchasing System focused exclusively on automation is going live very soon; for startups to make headway here, it’ll be an essential framework to be on. Unlike Spark DPS – it won’t be a startup-SME exclusive framework, but make sure to get on it early to stay in the competition.  

In terms of the opportunities available, there’s a lot of different problems AI can solve. From conversational AI and chatbots to strategic data analysis and operational decision making, the possibilities are (almost) endless. However, we always recommend that it’s best to understand what the issues that the authority is having are – take time to engage with them and their issue and see whether your solution is right for their need.

Finally, know that competition for the big contracts is still hot – IBM, Capgemini and others have won many of the most valuable contracts and will continue to do so, but the data shows that startups and SMEs with suitable solutions are at least as likely to win the most exciting contracts in this space. 

If you are a tech company that wants unparalleled insights into the key government opportunities and trends, get in touch with us about how we can help – you might be interested in applying to our GovStart programme

Using Tussell’s comprehensive database of UK public sector procurement and spend, you can get trends, analysis and live updates on the government’s procurement of technology. Essential for anyone looking to take a strategic approach to scaling up through public procurement.