Can the UK’s AI industrial strategy deliver?

There are three big assumptions at the heart of the UK government’s industrial strategy. Sir Keir Starmer’s administration thinks it can generate economic growth by making the public sector more productive, it hopes to stimulate construction and infrastructure builds, and it plans to achieve this over the course of this Parliament. 

On the face of it, it sounds like straight-forward policy-marking. That is until you look at the country’s poor record on output per hour – the typical economic measurement for productivity – over the last five years (link). The UK looks increasingly unproductive, with more and more taxpayer cash having to be spent on the welfare state as people exit the workforce. 

The rise of generative AI (from 2023 onwards) gives Whitehall an opportunity to break this cycle. But with tight fiscal rules and historic debt to GDP ratios, the government has a very limited budget to play with. It is within this framework that we should look at Starmer’s industrial strategy and its relationship to AI. 

A popular intervention from the government has been its pro-data centre stance. The UK will need to beef-up its own processing capacity if it wants to compete on the global stage and attract top AI researchers, who will want access to the best supercomputers. Starmer’s administration has therefore promised to make “timely planning decisions”, with a 13-week target for decisions made by ministers on called-in applications. 

This is part of Starmer’s wider commitment for the UK to be an “AI maker, not taker”. Though the government did pull £1.3bn of funding for a new supercomputer (link), the good news is the rather large tailwind from Silicon Valley, where the ‘Magnificent 7’ mega-cap technology stocks (Microsoft, NVIDIA et al) will spend twice as much on capital expenditure in 2025 compared with before the roll-out of ChatGPT. 

The UK hopes to grab a slice of that substantial pie, estimated to be more than $300 billion. Microsoft, as just one example, has committed more than $3 billion to expand its data centres in the country (link).

Talking of top talent, the new industrial strategy has promised to attract leading academics to the UK. It has launched a ‘Global Talent Taskforce’ and a £54 million talent scheme. The initiative will cover relocation and research costs over five years and its budget will be distributed by the UK Research and Innovation department to leading universities and other research organisations. 

On a related front, the government has also announced an AI skills drive, where it will work with top IT consultancies and AI providers to produce courses to help train up 7.5 million workers. The government has claimed that Whitehall is missing out on £45 billion per year of unrealised savings and productivity benefits, as 4-7% of public sector spend could be achieved through full potential digitisation of public sector services (link).

To help create these savings, ‘Humphrey’, the government’s own AI tool, has started to be rolled out amongst civil servants. “No one should be wasting time on something AI can do quicker and better, let alone wasting millions of taxpayer pounds on outsourcing such work to contractors,” Technology Secretary Peter Kyle has said. 

But beyond the public sector, the government’s interventions towards the private sector have been relatively light. There are no direct grants available for the adoption of LLM technology, for example, though R&D tax credits are still available for businesses. 

This could well be a missed opportunity. A recent Morgan Stanley paper (link) predicted that profits could grow by more than 16% thanks to the adoption of AI. But the jury is currently out as to whether the UK’s SMEs, which make up the backbone of the British economy, will quickly embrace the technology. 

Another potential road bump has been the decision from Matt Clifford, the Prime Minister’s top AI adviser, to step-down from government. Clifford is well respected within technology and finance circles. It’s unclear who will fill his role. 

Ultimately, if the AI industrial strategy is a success, the government’s finite bets will have to come in, while macro and micro-economic conditions outside of its control, including international investment, will have to fall in its favour.