The decision to scrap the Department for Science, Innovation and Technology (DSIT) has understandably caused alarm across the UK’s technology sector.
At a time when AI is threatening to reshape economies, national security and public services, dismantling the department responsible for technology can look like a step backwards. Critics are right to question whether the restructure will weaken specialist expertise or interrupt existing programmes, but there’s another way to view the decision. Burnham is right that technology and innovation should no longer be treated as separate sectors.
As AI becomes more integrated in our workplaces and everyday lives, it has outgrown the remit of one technology department. It now affects jobs, skills, investment, public services and how the government itself operates.
The same is true in business. AI can no longer sit on the edge of an organisation as a project owned solely by the technology team. Those scaling the value of AI are those seeing it more as a catalyst for wholesale transformation rather than an outcome in and of itself. The Government would do well to see it the same way.
AI cannot sit at the edge of an organisation
The government’s restructure reflects a challenge many businesses have been slow to recognise. AI has implications far beyond technology, so keeping it within one specialist team can make it easy for everyone else to treat it as someone else’s responsibility, limiting its full potential.
But spreading that responsibility is not enough on its own. Every part of an organisation needs to understand the part AI can (or can’t!) play in delivering the outcomes it cares about, whether that is improving customer service, reducing admin or making faster decisions. Otherwise, teams end up testing tools without a clear idea of what success should look like.
That is where many organisations are getting stuck. Businesses are investing in AI, but too few are turning those early experiments into wider, meaningful, sustainable change.
Britain is already stuck in AI pilot purgatory
Our recent research found that 42.5% of UK businesses are still experimenting with AI, while only 15.5% have successfully scaled it to create value. This is called the “AI pilot purgatory”. Businesses test tools, launch small trials and create innovation teams, but struggle to make AI part of how the wider organisation works.
In these instances, technology is rarely the main problem. Projects get stuck because ownership is unclear, leaders have not agreed what they want to achieve, or existing processes make it difficult to turn a successful trial into wider change.
Businesses need to be clearer about the outcomes they care about, their relative priority, who is responsible for delivering them and what happens when a pilot shows promise.
There also needs to be enough coordination at the centre to stop different teams repeating the same work, buying similar tools or solving the same problems in isolation. Wider collaboration can help AI become part of the organisation, but only if it is matched by clear accountability.
That is the same challenge the government now faces. Spreading responsibility for AI could help bring it closer to the decisions that matter, but without clear ownership it risks creating more pilots without more progress.
Too many leaders still ask: “Where can we use AI?”
Clear ownership is only part of the answer. Leaders also need to be much clearer about what they want AI to achieve, and be clear why AI is the right tool for the job, especially given the cost of running AI is less than certain going forwards.
Too many organisations still start by asking: “Where can we use AI?” It is an understandable question, but it often leads teams towards the most visible use cases rather than the most valuable problems. They buy tools, automate individual tasks and measure how many employees have adopted a platform.
A better starting point is: “What are we trying to achieve, and what needs to change to make that possible?”
For a business, that might mean reducing the time teams spend on admin, improving customer service or helping employees make faster decisions. AI may be part of the answer, but so will changes to processes, data, skills and accountability.
The same principle applies to the government. Rather than starting with where AI can be deployed, departments should start with the outcomes they are responsible for improving, whether that is reducing NHS waiting times, speeding up planning decisions or identifying fraud earlier.
This is why moving AI closer to the centre of government could matter. It creates an opportunity to make AI part of broader conversations around productivity, jobs, investment and public services, rather than treating it as a separate technology programme.
Judge the decision by what happens next
Scrapping DSIT isn’t automatically reckless despite what the critics say. Its success will depend on whether the new structure brings AI into the decisions that matter. It must retain specialist scientific and technical capability while giving other departments genuine responsibility for applying it.
Businesses can learn from what comes next. Removing a department or changing who owns AI is the easy part. The harder question is whether that change helps AI move beyond individual pilots and become part of how the organisation works.
For businesses, that means looking at whether different teams have clear responsibility for applying AI, whether successful experiments have a route to scale and whether investment is tied to a clear business outcome. Simply moving AI out of the technology team will not solve the problem if the rest of the organisation is not set up to take ownership of it.
The same will be true of the government’s restructure. The real test is whether it can turn ambition into adoption, move experiments into scaled services and turn technology investment into better outcomes for the public.
