Edinburgh study shows Scotland leads the UK in local government AI adoption
New report highlights rapid progress in public‑sector AI
Scotland is emerging as the UK’s leading nation for local government AI adoption, according to a new comprehensive study carried out into how councils are using the technology.
The 18‑month study, led by Heriot-Watt University and GoLLM examined how ready local authorities are to adopt and use different AI applications by analysing two hundred and eight councils across Northern Ireland, Wales, Scotland and England.
The report, titled AI Readiness of UK Local Authorities 2025, shows Scotland pulling ahead across multiple fronts, with nearly one in three councils now using or testing AI.
The research was carried out by Heriot‑Watt University’s Intelligent Automation Systems Lab (IAS), AI specialists GoLLM and made possible through partnership with the Association for Public Service Excellence (APSE), the Society for Innovation, Technology and Modernisation (Socitm), Coram Children’s charity and Sustainable Scotland Network (SSN).
Dr Luciana Blaha, IAS lab lead at Heriot-Watt University’s Edinburgh Business School said: “Scotland’s progress is being driven by a coordinated national approach, strong collaboration across the public sector and close partnerships with universities.
“National programmes such as the Scottish AI Strategy, the Scottish AI Alliance, the Digital Office and COSLA’s governance frameworks have given councils the confidence and structure they need to move forward.
“Scotland’s coordinated approach and strong partnerships put it in a strong position to be a key node the next phase of public‑sector AI adoption, setting an example for the rest of the UK.”
Funding of over £7k through the Innovation Voucher Programme, managed by Interface, which links organisations to academic expertise, and £3k from the IAS Lab supported the study, and additional in‑kind support from GoLLM enabled a team of eight researchers to complete the work.
In addition, GoLLM has since invested more than £6k in research analysts to continue supporting the project following the conclusion of the Interface and IAS Lab funding.
Dr Blaha added: “The wider UK findings show a sector keen to adopt AI but held back by uneven data foundations, culture and limited capacity.
“We found that most councils are still in a difficult middle ground, where they recognise the potential of AI but lack the data foundations, confidence and governance strategy to use it safely and at scale.
“There is huge variation between councils, often shaped by leadership priorities, legacy systems and the speed at which clear guidance is developing, which makes it much harder for many to move beyond early experimentation.
“Despite this, we saw encouraging examples where AI was introduced in ways that strengthened services and supported staff rather than replaced them, which shows responsible, community‑focused, and sustainable adoption initiatives which differ by region.”
The report found uses of AI in Scotland ranging from improving waste services to streamlining customer support.
In Fife, AI is being used to help sort household recycling, while Glasgow is developing a new materials recovery facility using advanced robotics and optical technology.
Aberdeen has taken a different approach, using AI in its internal administration through its AB1 chatbot, which supports residents around the clock in multiple languages, showing a good example of cultural awareness, such as through its Doric vocabulary ‑1 chatbot, which supports residents around the clock in multiple languages.
Orkney is also highlighted as a strong example of how smaller councils can adopt innovative tools effectively when they have a clear community focus, leading on integration of AI in small pilots, as well as planning towards larger projects, including digital twins.
Daniel Shorr, CEO of GoLLM and co‑author of the report, said “This research turns a complicated landscape into a clear starting point for action.
“It shows that AI readiness isn’t determined by size alone, the councils pulling ahead are distinguished by leadership ambition, governance discipline, strategic clarity and stronger data capability.
“Those are the same foundations we design for in our feedback-to-insights platform D.A.V.E. (Dynamic Analytics Visualisation Engine) which streamlines public consultations and other feedback data into actionable reports.”
The report says Scotland can stay ahead by growing shared digital systems, helping councils work more closely together, and improving training so staff can build stronger data and AI skills.
It also encourages councils to work more regularly with partners such as APSE, Socitm and universities to share what’s working, exchange feedback and build confidence across the sector.
The findings come as councils face rising demand for services and increasing pressure to modernise systems while protecting frontline staff time.
A spokesperson for APSE said: "APSE was delighted to support the “AI Readiness of UK Local Authorities 2025” report. The findings reflect what we hear from our members regularly: councils are keen to innovate but progress depends on strong governance and organisational capacity.
“This research provides a timely and practical look at where local government stands on AI, and what’s needed to move from pilots to meaningful and scalable impact for communities.”
A spokesperson for Socitm added: “Socitm's own research aligns with the findings from this study, which highlight that enthusiasm for AI is growing across local government, but many councils are still navigating practical challenges, from legacy systems and siloed data to digital skills gaps.
“Working together with Heriot-Watt, GoLLM, and other sector partners we want to help councils overcome these barriers by strengthening data foundations, building capability, and embedding ethical governance, so they can turn ambition into trusted, tangible improvements in public services.”
Professor Luciana Blaha and the research team will share further insights at the AI Public Sector week on 17 March, and for an APSE member-only session on 24 March.