There is no escaping the noise around artificial intelligence. Yet behind the breathless launches and boardroom enthusiasm sits a far more sober question, and it is one MPs are now determined to answer: are British businesses, and the workers inside them, actually getting anything out of it?

That question has become harder to dodge over the past two years. Having consulted hundreds of firms the length and breadth of the country, the Business and Trade Committee (BTC) has heard a recurring worry, that the UK is trailing competitor nations when it comes to helping companies, and small firms in particular, put AI to work. The risk is not merely missed efficiency gains. It is the prospect of British business losing the race for competitive edge before it has properly begun.

The flip side is just as instructive. A steady drip of embarrassing headlines, professional consultancies serving up error-strewn reports stuffed with invented citations and references that simply do not exist, has exposed the perils awaiting the unwary early adopter. For every firm quietly banking the benefits, another is discovering that AI without judgement is a liability dressed up as a shortcut.

It is against this backdrop, and as the Government presses ahead with a fresh raft of measures to spur development, uptake and use of AI, that the committee has opened its inquiry into artificial intelligence, business and the future of the workforce. Over the coming months it will test the attitudes and approaches of big and small business alike, alongside the public sector.

The terms of reference are refreshingly blunt. What real, here-and-now benefits is AI delivering to British business, the health service and local government? Which gains remain stubbornly out of reach, and what is blocking them? Is adoption in some sectors leaning too heavily on a handful of large technology platforms? And what should we make of the curious finding that Britain’s micro-businesses appear to be embracing AI more readily than their small and mid-sized counterparts?

Above all: what do we stand to lose?

The prize is not trivial. The OECD has estimated that AI adoption could add between 0.4 and 1.3 percentage points to UK productivity growth, worth tens of billions in additional output by the end of the decade. Realising even a fraction of that would move the dial on a productivity problem that has dogged the British economy for the best part of two decades.

Yet the evidence already gathering on the desks of the nation’s business press suggests the gains are real but uneven. Smaller firms are reporting quick, low-cost productivity wins, drafting copy, planning staff rotas, trimming waste, handling routine customer queries, long before they tackle anything more ambitious. The tools are cheap and, for the most part, straightforward to deploy. The harder, more valuable transformations remain the preserve of the few.

The committee’s interest in firm size cuts to the heart of the matter. Adoption is not spread evenly across the economy, and the reasons are familiar to anyone who has watched smaller firms wrestle with new technology: thin margins, scarce digital skills, a shortage of time to experiment and a justified wariness of betting the business on an unproven tool. Closing that gap, and unlocking the growth that AI promises UK SMEs, is rapidly becoming the defining test of whether the technology delivers for the whole economy or merely the well-resourced top of it.

Rt Hon Liam Byrne MP, Chair of the committee, framed the challenge in characteristically direct terms. “We can all see the excitement around artificial intelligence, but what is less clear is whether enough British businesses are actually using it to improve productivity, cut costs and win new customers,” he said. “We have heard growing concerns that while some firms are racing ahead, too many others, especially smaller businesses, are struggling to adopt these technologies at scale. If that is true, Britain risks falling behind competitors who are moving faster.”

He was equally alive to the downside. “At the same time, there are obvious questions about reliability, security and trust. Stories of AI systems producing flawed analysis, fabricated references and poor advice underline the importance of getting this right. Our inquiry will examine where AI is genuinely making a difference, what is holding back wider adoption, and what government and industry must do to ensure the benefits are spread across the economy. The challenge now is not just to invent the future, but to make sure Britain is equipped to maximise it.”

That last line bears repeating, because it captures the whole exercise. Invention has never been Britain’s weakness. Diffusion is, getting good ideas out of the laboratory, off the conference stage and onto the shop floors and back offices of the country’s 5.5 million businesses. On that score, the jury is still very much out, and the committee’s inquiry could not be better timed.