Summary
From December 2024 to February 2025, the Local Government Association (LGA) repeated its survey to explore the use of Artificial Intelligence (AI) in English councils. The purpose was to build an updated picture of where AI is currently being deployed in local services and council business units and to map where the greatest opportunities and risks lie. This will enable the LGA to build an evidence base for its support to councils in this space, and to ensure that local government is part of the national conversation.
The survey used the Government’s definition of AI:
The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages. Modern AI is usually built using machine learning algorithms. The algorithms find complex patterns in data which can be used to form rules."
November 2023, Introducing the AI Safety Institute.
It also defined four types of AI, based on information provided by the Alan Turing Institute:
- Perceptive AI, such as systems that recognise faces and fingerprints, or try and analyse images, audio or video, for example in the analysis of consultation responses or identifying car registration plates in the prevention of fly tipping. This includes sensing AI such as remote or continuous sensing through smart sensors.
- Predictive AI, such as systems that try and make a prediction about an outcome for an individual, or try and assign people to appropriate service or system, for example predicting an outcome in services or assigning an adult social care treatment pathway.
- Generative AI, such as systems that generate text or images, such as ChatGPT and DALL:E.
- Simulation AI, such as digital twins and agent based modelling.
Key findings
Responses were received from a third of councils (33 per cent), as such, the results of the survey should not be taken to be more widely representative of the views of all councils. Rather, they are a snapshot of the views of this particular group of respondents.
Almost all respondents (95 per cent) were using or exploring AI with half at the beginning of their AI journey, 22 per cent developing their AI capacity and capabilities around AI, 14 per cent making some use of AI while 7 per cent are innovative and considered as leaders among councils in their use of AI. This indicates that councils are progressing in this area, as the overall proportion using or exploring has increased by 10 per cent since 2024.
Generative AI was the most commonly adopted type being used or explored by respondents (83 per cent). This was followed by perceptive AI, which had been adopted by 28 per cent and predictive AI, (systems that try to make a prediction about an outcome) which was being used by 20 per cent. Changes in relation to the types of AI being adopted also showed that councils were moving forward in their AI journey, as the proportion who reported that they weren’t using it decreased by 8 per cent.
As with the original survey, the functions where respondents using or exploring AI had most commonly utilised it were corporate council use: HR, administration (meeting minutes), procurement, finance, cyber security (84 per cent), health and social care (adults) (44 per cent) and health and social care (children’s) (31 per cent).
Just over two-thirds (68 per cent) of the respondents using or exploring AI were paying external suppliers for the provision of AI tools or technologies, or were in the 6 process of procuring this, a slight increase on the 63 per cent found in the original survey.
Among respondents who were using or procuring external suppliers for the provision of AI tools or technologies, two-thirds (66 per cent) identified ‘project scoping: understanding where AI can add value’ as representing a barrier to a great or moderate extent, this was followed by scoping requirements: understanding how AI is embedded in a product (61 per cent) and ‘evaluation: understanding how to evaluate solutions’ (58 per cent). The proportion of respondents who identified these as barriers was lower than in the original survey, suggesting growth in their levels of understanding.
The areas where most respondents had realised benefits from using AI were staff productivity (36 per cent), service efficiencies (33 per cent) and cost savings (21 per cent). The areas where respondents saw the greatest AI opportunities were corporate council use: HR, administration (meeting minutes), procurement, finance, cyber security, identified by 88 per cent, followed by health and social care (adults) (49 per cent), and advice and benefits (38 per cent). These areas were the same in the original survey.
In common with the original survey, the three biggest barriers to deploying AI identified by respondents were lack of funding (62 per cent), lack of staff capabilities (56 per cent) and lack of staff capacity (52 per cent).
The issues most commonly considered to represent a great or moderate AI risk were cyber security (83 per cent), deep fakes disinformation (69 per cent) and organisational reputation and resident trust (68 per cent). There were the same three most commonly cited risks as in the original survey.
Half (49 per cent) of respondents were using existing policies to manage AI risk, 41 per cent used a specific AI policy and 38 per cent used a Senior Responsible Owner. These were slightly different to the original survey due to a 13 per cent increase in those using a specific AI policy.