The UK‘s aspirations to modernize its public sector with AI may face serious obstacles, according to a new report from the Public Accounts Committee (PAC). The report highlights the poor quality of government data and outdated technology systems as significant barriers to the effective use of AI.
For AI to function optimally, it requires high-quality data, but the report reveals that much of the government’s data is fragmented and stored in legacy systems that are no longer supported or impossible to update.
As of last year, around 28% of central government systems were classified as legacy. There is no funding allocated to modernize 21 of the 72 highest-risk systems. This issue casts doubt on Prime Minister Keir Starmer’s recent pledge to make the UK a global leader in AI, which includes promoting AI infrastructure in the public sector.
The report also discusses the government’s efforts to address AI-related challenges, such as the expansion of the Department of Science, Innovation and Technology (DSIT) to oversee AI initiatives. However, these plans seem unlikely to succeed without addressing fundamental issues like data fragmentation and poor IT infrastructure.
Moreover, the PAC highlights a severe skills shortage, with many digital and data roles in the civil service going unfilled. The report warns that the government is moving too slowly in terms of transparency and could become overly reliant on large tech companies, limiting innovation. Sir Geoffrey Clifton-Brown, chair of the PAC, expressed concerns that the government’s digital strategy lacks the necessary authority and vision for a real transformation.
The report ultimately casts doubt on whether the government can modernize its systems effectively, suggesting that efforts to bring the public sector into the 21st century may face insurmountable challenges.