By Marc Stephenson |

April 11, 2025

Government AI roll-outs threatened by outdated IT systems, poor data and skills shortage

According to the State of Digital Government Review, published on 21 January this year, the UK public sector spends over £26 billion annually on digital technology, employs a workforce of nearly 100,000 digital and data professionals, and delivers millions of online transactions every day.

And the government has more ambitious plans to drive growth and boost efficiencies by embedding AI in all aspects of its work, which will deliver billions of pounds in savings to the government. However, a cross-party commons committee, the Public Accounts Committee (PAC) has warned in a recent report that there are several challenges to address in order to successfully introduce AI initiatives in the public sector.

Public Sector AI

AI adoption in Public Sector: Key challenges

First and foremost, the Public Accounts Committee reveals that out-of-date technology is a major concern. The report by the PAC found that more than 20 central government IT systems have been identified as “legacy”, meaning out of date and unsupported, and have yet to be given funding to improve them. In fact, legacy technology is currently estimated to represent almost a third of systems in central government departments in 2024. And alarmingly, 47% of services offered by central government still rely on non-digital methods like phone calls and paper forms.

Secondly, there is an issue around poor quality data. Most public sector organisations are independent, choosing to build and maintain their own technology estate, inhibiting standardisation, interoperability and reuse. This is reflected in the State of Digital Government Review, which reveals that public sector data is fragmented and underused both within and across organisations, which will subsequently hold back AI, machine learning, and advanced analytics. Almost three quarters (70%) of survey respondents for the review admitted their data landscape is not well co-ordinated, interoperable, and does not provide a unified source of truth.

In a speech earlier this month, Starmer said AI should replace the work of government officials where it can be done to the same standard, with 2,000 new tech apprentices to be recruited to the civil service. However, the PAC report warns about digital skills shortages in the public sector, in part because of civil service pay levels that are uncompetitive with the private sector.

This is the third challenge. Effectively, the public sector struggles attract and retain top digital and data talent. It is reported that 50% of digital and data recruitment campaigns failed in 2024, which represents a decline in performance from 22% in 2019. The State of Digital Government review goes on to share that there is typically a 35% pay gap between public and private sector digital roles – equivalent to £30,000 per year.

Other root causes raised in the State of Digital Government review address a lack of funding for digital and data projects as well as a lack of leadership in central government to drive digital initiatives

Transform and reform

The targeted use of AI has real potential for large-scale public sector benefit: from introducing tools to automate the drafting of policy documentation to reduce administrative burden, to implementing voice automation to reduce call costs and improving user experiences. However, successful use cases remain few and far between.

It is clear that the public sector is not ready to introduce AI at the speed suggested by government. The report has highlighted key areas that need to be addressed, including leadership, structure, talent and funding, to transform and reform the way government does digital and the way we, as citizens, interact with the state.

Let’s talk

We work with many central and local government agencies, providing a variety of services such as information architecture, records management, and information governance, as well as AI services, including readiness assessments and deployment. All these services are featured on G-Cloud 14 – you can find Metataxis as a G-Cloud 14 Lot 3 supplier for Cloud Support.

Find out more about our AI services or simply get in touch for a non-obligatory chat about your AI plans.

Discover how we can help you adopt, deploy and embrace AI.