In a quiet office on the outskirts of Manchester, a mid-sized manufacturer quietly integrated text-generating tools into its daily reporting last year. The change shaved hours off routine tasks. No government scheme prompted it. No regulator demanded it. The decision came from managers who saw a practical edge and acted.
That small scene mirrors a broader pattern released by the Office for National Statistics on 2 July. Nearly three in ten UK businesses reported using at least one type of artificial intelligence technology in June. The figure stands eight percentage points higher than the same period a year earlier. Among companies with 250 or more employees, adoption reached 49 per cent, up 13 points from the previous June.
The data, drawn from Wave 159 of the Business Insights and Conditions Survey conducted between 15 and 28 June, captures a steady, unspectacular climb. Text generation using large language models led the way at 17 per cent of businesses, followed by visual content creation at 14 per cent. Both categories have risen sharply since the questions first appeared in September 2023, climbing 12 and 11 points respectively.
Larger firms set the pace
The gap between small operations and their bigger counterparts reveals something fundamental about how technology spreads. Larger businesses, with more resources to experiment and absorb risk, move faster. They test, discard what fails, and scale what works. Smaller firms follow when the value becomes obvious and the tools cheaper. This organic sequence has powered every major productivity shift in modern economies.
Yet the modest pace also carries a warning. Each additional layer of regulation, each new compliance cost, each uncertainty about future rules risks flattening the curve that has so far risen through market signals alone. When governments insert themselves too deeply into decisions about which technologies deliver real returns, they distort the very incentives that drive adoption. The businesses closest to customers and costs know best where artificial intelligence adds value and where it does not.
The survey offers no fresh detail on barriers this time. Earlier waves pointed to skills gaps, upfront expenses and regulatory uncertainty as common obstacles. Those factors remain real. But the steady upward trend, even amid pressures from energy prices, international tensions and shifting economic conditions, suggests many UK enterprises are cutting through the noise. They adopt where the benefit is tangible and ignore the hype where it is not.
