Unlocking the benefits of AI adoption in the UK public sector: the evolving landscape

Why should I read this?

This briefing note spotlights recent Artificial Intelligence (AI)-focussed private sector partnerships entered into by the UK government, evolving policy initiatives and their wider impact on public sector customers and suppliers.

Adoption of AI by the public sector and more widely is viewed by the UK government as a significant lever to delivering economic growth. The AI Opportunities Action Plan published at the start of the year aims to “shape the AI revolution” in the UK by:

• investing in AI infrastructure, talent and regulation;
• driving AI adoption across various sectors of the UK economy by piloting and scaling AI products; and
• positioning the UK market to be at the forefront of developing frontier AI-powered solutions.

A live procurement tracker (as of August 2025) reports that between January 2018 and end of July 2025, the UK public had awarded AI contracts worth £3.45 billion, demonstrating that the total value of AI contracts had grown by a staggering 887% between 2018 and 2025.

Public-private AI partnerships

A snapshot of recent notable AI initiatives undertaken and partnerships entered into the by the UK government is discussed below. This trend has in turn prompted technology suppliers to set up dedicated public sector divisions within their businesses to unlock the commercial benefits of the growth in AI driven public procurement opportunities.

• In July 2025, the UK Department for Science, Innovation and Technology (DSIT) has entered into a strategic partnership agreement with Google to launch a training programme for 100,000 civil servants working across the public sector to improve their understanding and usage of emerging technologies such as AI and other digital services by 2030, to support:
• the UK government’s goal to have one in ten civil servants working in technology roles by the same date and to modernise the public sector;
• drive greater collaboration between industry and DSIT AI developers;
• reduce the UK public sector’s reliance on legacy technology by replacing aging and outdated IT systems with more agile, secure, cloud-based technologies and providing technical support to central and local government customers;
• create a cross-government cyber security platform that will monitor and respond to cyber security threats; and
• wider economic growth via improved public services and delivery of greater value for taxpayers.
• The collaboration with Google has additionally led to the creation of a specialist AI tool called ”Extract” which is designed to assist local councils in digitising handwritten planning documents and maps in minutes.
• Other leading technology firms like Amazon Web Services have committed to training 100,000 UK students in AI, cyber security and data skills by 2030 as part of its ‘skills to jobs tech alliance’.
• Meta has earlier this year committed a $1 million grant to the Alan Turing Institute for an Open-Source AI fellowship which will fund:
• ten fellows who will work with the UK government for one year to build AI tools for “high-security use cases” in the public sector, such as language translation for national security, using construction data to speed up approval processes/unblock planning delays and reduce the cost of integrating AI across various government functions; and
• the development of “Humphrey”, a suite of AI-powered tools for civil servants to help them effectively deliver on ministerial requests.
• In July 2025, OpenAI and the UK government signed a new strategic partnership setting out plans to expand AI security research collaborations, explore investment in UK AI infrastructure like data centres and identify innovative ways for provision of taxpayer funded services like security and education.
• The UK government is also testing various AI tools for the efficient and cost-effective delivery of public services such as  “Caddy”, an open-source AI assistant used at Citizen’s Advice centres which gives the users of government call services advice on queries regarding debt management, citizens’ rights and legal support.

National agentic AI tender

Additionally, a pre-market engagement was launched earlier this year in March 2025 by DSIT in respect of a National AI Tender for a phased engagement with frontier AI companies to support agentic AI powered enhancement of existing GOV.UK  services. The ‘National AI Tender’ approach follows a ‘Scan, Pilot, Scale’ framework (as also outlined in the above mentioned AI Opportunities Action Plan), with the goal of setting up a framework for AI powered national procurements to deliver a unified and citizen-centric digital government platform. Specifically:

• the ‘Scan’ phase covered two-way partnerships with AI suppliers to understand current and future agentic AI technology capabilities to develop personalised guides to reduce the complexity faced by users in accessing public services across different government departments, alongside recognising the nuances of individual life situations and experiences;
• the ‘Pilot’ phase, set to be undertaken in 2025/26, will focus on testing the value, technical feasibility and deliverability of using agentic AI powered systems for key life transitions such as education pathways, apprenticeship opportunities, and career guidance for 16-34 year old citizens; and
• the ‘Scale’ phase, intended for 2026/27, will focus on rolling out a fully integrated, agentic AI powered nationwide GOV.UK chat service to all user groups.

In light of the emerging AI adoption trends by the public sector in the UK, it is critical for public sector customers and suppliers to prioritise the recognition, identification, and assessment of commercial, legal and compliance risks when procuring and developing AI powered IT solutions.

Empowering public sector AI literacy and impact on public procurements

The AI Playbook published in February 2025 sets out clear practical guidance and ten operational principles for civil servants on the safe and effective use of AI including: (i) learning the essentials of AI; (ii) applying AI daily in their work; and (iii) improving public service delivery by integrating AI into their professional toolkit.

The AI playbook also points towards existing detailed guidance and other government playbooks to facilitate public procurement of AI including the following: 

• UK Treasury’s Green Book which specifies guidance for developing a robust business case for public sector procurements;
• the Digital, Data and Technology Playbook which specifies the key policy reforms for how the UK government should assess, procure and deliver digital projects, which all central government departments and their arms-length bodies are expected to follow on a ‘comply or explain’ basis;
• the Sourcing Playbook which specifies key policy, best practices and guidance for making sourcing decisions for the delivery of public services; and
• specific Procurement Policy Notes (PPNs) such as “PPN 02/24: Improving Transparency of Artificial Intelligence Use in Procurement” which sets out optional questions to help identify the use of AI in public procurements and manage risks via disclosure and due diligence requirements.

Additionally, AI literacy and adoption is being further promoted via the UK government’s ‘One Big Thing’ initiative which is focussed on training and upskilling civil servants on AI. Notably, the recent Alan Turing Institute report, titled ‘Mapping the Potential: Generative AI and Public Sector Work’ published in June 2025 analyses how generative AI could enhance productivity in the UK public sector and unlock unrealised productivity benefits via digital infrastructure modernisation and by strategically mitigating the impact of AI adoption barriers. Key trends and insights identified in that report include:

• AI-driven time savings: on average, 41% of public sector work time could be supported by generative AI (i.e. circa 3.5 hours of an 8-hour workday); and
• Tailoring: the report recommends that AI strategies should be tailored, as opposed to a one-size-fits-all approach, such as customising AI adoption based on sector, role (e.g. frontline versus non-frontline) and task type.

The report also emphasises the significance of promoting AI literacy alongside the recognising the essential need to retain human oversight where tasks involve an application of discretion or judgement and interpersonal interactions in the delivery of public services. Standard IT service contracts catering to legacy technology may not adequately address the unique risks associated with delivery of AI powered services and systems and should ideally be bolstered with well-drafted and comprehensive IT contract terms and conditions which are specifically customised for AI usage.

Concluding remarks

The upward trajectory in AI adoption by the public sector, promotion of AI literacy including in the context of public procurements and the evolution of the underlying policy landscape is likely to continue, as the UK government continues to seek opportunities at pace to unlock AI powered productivity benefits and to make the provision of the public services more efficient.

Correspondingly, public sector customers and suppliers alike should:

• recognise and adapt to the new era of AI powered technology transformation, alongside familiarising themselves with the capabilities, limitations and risks of procuring and deploying tailored AI solutions in the provision of public services; and
• ensuring that the commercial terms on which they contract not only keep pace with the evolving policy landscape but are also future-proofed.

For more information

Eversheds Sutherland’s Technology Team has extensive experience in advising public and private sector clients on cutting edge, complex and high value IT transformation projects using various contracting approaches including the Crown Commercial Services’ Model Services Contract and delivering training on the implications of the new public sector procurement regime ushered in by the Procurement Act 2023. Please contact Nick Blane, Madhulika Kanaujia, Joanne Veitch or Elizabeth Marshall if you would like to discuss in more detail.