Slow uptake of AI in government hindering strategic goals, new research finds

Enthusiasm for adopting AI to improve public services is hindered by practical hurdles to implementation

By SAS

09 Oct 2024

While the UK government has recognised the transformative potential of AI, outdated IT infrastructure, budget constraints, skills gaps, operational procedures, and data issues are challenging its adoption and implementation. These were the findings in a new research report Data & AI in the UK government: Overcoming barriers and realising potential from data and AI expert SAS, conducted in partnership with Civil Service World’s sister organisation, Total Research. The study follows on from similar research in 2022, surveying 170 officials from across government about the current state of digital transformation.

Only around one in five respondents report ‘a lot’ or ‘significant’ progress in key AI-related areas, such as automation technologies (22%), machine learning/artificial intelligence (17%), and generative AI (12%).

‘Cost and budget restrictions’ are the primary concerns for the lack of progress among two-thirds of respondents (67%), followed by a ‘lack of internal technical knowledge or expertise’ (63%). More than half cite outdated legacy systems (57%) as a barrier.

Meanwhile, concerns about data security, privacy, and fragmented systems hinder progress for 52% of respondents. Data quality and consistency issues are also prevalent, affecting more than 40% of respondents who cite non-standard records (45%) and the ‘availability of high-quality data’ (42%) as significant problems.

The slow uptake underscores a gap between strategic goals and practical implementation, following the government’s new vision for digital and data.

The Government Digital Service (GDS), Central Digital and Data Office (CDDO), and the Artificial Intelligence Incubator have been relocated to the Department for Science, Innovation and Technology (DSIT) to enhance digital services and innovation in both the public and private sectors. In line with this goal, when asked about their priorities, 22% of respondents ranked ‘harnessing AI-driven capabilities for automation', up from 12% in 2022.

Nicola Furlong, Director for Public Sector, Health and Life Sciences at SAS, commented: “Enhancing efficiency and improving the services to citizens are just two of the benefits AI can offer, but the integration of this technology in the public sector is lagging behind. It’s clear that issues surrounding data privacy, security, and ethical AI are influencing how these technologies are implemented.

“Ensuring robust governance and adherence to ethical standards is critical to the successful adoption of AI and GenAI, in addition to substantial investment in infrastructure and expertise to effectively integrate GenAI into existing systems. Government departments should begin by planning and allocating resources, then identifying use cases before developing AI prototypes, followed by implementation and ongoing assessment to maintain the solution.”

The government's approach to data usage is captured in the Functional Data Standard (FDS), which is designed to ensure organisations adhere to rigorous guidelines, creating a foundation for robust digital and data initiatives.

The survey was conducted when the FDS was not officially launched. Yet, it’s encouraging that 28% of civil servants are aware of it, 14% have actually seen it, and 3% are actively reviewing it.

Patrick Haston, Head of Cloud Transformation, Cabinet Office, said: “Compliance with the data standards enables data to be shared across government more easily, supports greater automation and provides opportunities to use AI more effectively. Surveys such as this provide valuable insight into levels of awareness and adoption.”

The report details the five practical steps to help government departments and public sector organisations unlock the full potential of data and AI, and is available to read at Data & AI in the UK government: Overcoming barriers and realising potential.

Download research report 

Read the most recent articles written by SAS - Building a case for a change: Tackling fraud with AI

Share this page