What DSIT’s breakup mean for the UK’s AI and digital ambitions?

Estimated Reading Time: 4-6 minutes.

In the run up to Burnham becoming PM, numerous voices in the tech sector sounded the alarm about plans to abolish the Department for Science, Innovation and Technology (DSIT) – a department that had made a name for itself pursuing the Labour government’s ambitious plans for AI both inside government and across the economy. Labour have rightly sought to reform the machinery of government in their ambition to ‘rewire the state’, and AI and digital innovations have been an integral part of that agenda.

Those fears were confirmed last month. Liz Kendall was dismissed as Secretary of State and the Department’s functions were divided across three departments. The Government Digital Service (GDS) has been moved to the expanded Department for Digital, Culture, Media and Sport (DCMS), AI now sits in the Cabinet Office under the new Minister for Artificial Intelligence Kanishka Narayan, and science and innovation has been moved to a revamped Department of Business, Innovation, Science and Trade (BIST).

The decision to disband DSIT has been roundly criticised owing the good reputation the department managed to build within its relatively short lifespan, with the Institute for Government (IfG) describing the creation of DSIT in 2023 as an ‘all-too-rare example of a machinery of government change that was thought through and had clear logic’. Since it was established, a number of Cabinet Office digital units, including GDS, were placed under its remit and tasked with the introduction of AI tools and systems across government and the economy. Given the importance of high-tech sectors in the Labour government’s growth plan, and the belief that AI could provide the ‘magic bullet’ for reforming government, there was a clear logic behind preserving the department and even uplifting its budget.  

A Pattern of Incoherence

How, then, should the redistribution of AI and digital competences under the Burnham government be appraised?

First, some of the changes would appear to give greater attention to the AI remit.

The positioning of the overall AI agenda in the Cabinet Office signals a return to an earlier and arguably proven model of digital transformation. The original Government Digital Service and the Central Digital and Data Office were led from the centre, as was the ground-breaking work of the Incubator for AI. The former two agencies have been widely acclaimed for breaking new ground in the digital modernisation of central government, including the successful establishment of the GOV.UK platform and improvements in cross-departmental data sharing, whilst the Incubator for AI pioneered new public sector AI tools, including the flagship ‘Humphrey AI’ suite.

Additional factors suggest that this is not a downgrading of AI ambitions, but rather a rescaling and rescoping of them. This includes the appointment of highly regarded Kanishka Narayan MP as Minister for Artificial Intelligence – a newly established position that grants him a place in the Cabinet.

AI has also been placed at the centre of the government with a new AI Taskforce, chaired by Lord Vallance. The Taskforce will oversee how AI can transform public services and economic growth and will report directly to Narayan and Cabinet Secretary Dame Antonio Romeo, who has made it clear that AI-driven transformation of the Civil Service is a key priority for her.

Furthermore, it would seem wise to combine DSIT’s commercial agenda with the newly expanded BIST, particularly in light of the ‘Modern Industrial Strategy’ and the key role played by AI and digital technologies within and across it. From tackling barriers to AI adoption in the private sector, to establishing an ambitious AI Hardware Plan, there was clear overlap between the two departments focus on how science and innovation could be leveraged to address longstanding productivity issues.

However, other aspects of DSIT’S breakup appear to reflect a wider pattern of incoherence within the UK state.

Most of all, moving GDS into DCMS does not seem to reflect its historic role and importance. GDS was pivotal in the UK government’s ascent to the top of the United Nations (2026) E-Government Development Index rankings in 2016, and this agenda has become yet more significant in the age of generative AI. By moving GDS sideways to a department without an exclusive digital mandate, the Burnham Government does not appear to be giving the digital transformation agenda the prominence or resources it warrants.

DSIT was also overseeing key initiatives that have now been left in limbo. Chief among these is the AI Growth Zones, whose progress has stalled and whose future is now in doubt. This was a major pillar of the Starmer government’s plan to ‘unleash AI’s potential’ to drive economic growth, but responsibility for the plans have not yet been reassigned. This is important, as AI Growth Zones were introduced as means of addressing some of the critical bottlenecks (such as grid availability) to AI data centre buildout. The uncertain future of AIGZs could thus further undermine already faltering investor confidence in the UK market. 

Additionally, DSIT was also in the process of expanding and improving access to the AI Research Resource, the UK’s primary publicly funded suite of supercomputers which provide compute to researchers, academia and industry. Whilst no one is suggesting the future of AIRR is in doubt, questions inevitably emerge regarding funding, institutional capacity and focus which may be threatened by a lengthy departmental restructure, changes in personnel, and shifts in priorities.

What does this signal about Burnham’s digital and AI agenda?

Overall, the diffusion of DSIT’s responsibilities without a clear overarching strategy raises the spectre of increasingly incoherent governance of AI and digital policy in the UK. Whilst some elements of the reorganisation warrant praise, issues such as institutional, personnel and policy churn across multiple poorly coordinated departments and public bodies, combined with a short-termist, often reactive approach to policymaking, have long contributed to a sense of fragmentation in UK governance.

The breakup of DSIT’s agenda across multiple departments risks exacerbating this sense of fragmentation and weakening coordination. It may also be seen to reflect Burnham’s apparently more sceptical view of the digital and AI agenda, as illustrated by the scrapping of the Digital ID scheme, his more cautious statements on AI and Big Tech, and reports that he is set to end the NHS’s contract with Palantir. 

And yet the elevated role of AI in central government appears consistent with the government's default tendency when it wants to signal importance of a particular policy problem. Transferring the AI remit to the Cabinet Office may, in fact, strengthen the centre’s capacity to coordinate AI policy across government. DSIT’s merger with BIST also suggests a clearer commercial focus for the government science and innovation policy.

Therefore, on ministers return from parliamentary recess, the main test for the Burnham government will be whether these new arrangements prove to be suitable for delivering a strategically coherent and accountable AI and digital agenda.

Darcy Luke is Research Associate at BRAID, University of Edinburgh, studying the social value of AI investment in the UK.

Bradley Ward is a Leverhulme Post-Doctoral Fellow studying AI policy under the Labour government at the University of Southampton. 

Nathan Critch is Research Associate at The Productivity Institute, University of Manchester, studying mission-led innovation policy.