TL;DR

A report released on 29 October by The Productivity Institute, based at Alliance Manchester Business School, finds that organisations are taking up AI well ahead of their ability to profit from it. Bytes Technology Group, the FTSE 250 technology company, commissioned The AI Ambition Gap, which was produced independently. Its central estimate is that AI’s most plausible contribution to UK productivity growth is between half a point and 0.7 of a point a year, provided firms put money into managers, training and new ways of working, not just tools.

Big gains that shrink

The report pulls together published studies and eight expert interviews rather than new survey work. It shows big task-level gains shrinking as they scale. Support staff with an AI assistant cleared 14% more queries per hour, and professionals completed writing and analysis 40% quicker. Programmers with AI assistants tripled their code output, yet released only 30% more software, because human checking, testing and release became the bottleneck.

Inside organisations, use stays shallow. Some 89% use AI somewhere, but most staff spend an hour or two a week on it at most. Only a quarter have redesigned work around it, 46% have got beyond pilots, and just 6% of firms earn at least 5% of profits from AI.

The authors, Bart van Ark and Vassilis Karountzos, also point to a “J-curve”: retraining, reorganisation and data clean-up cost money at once, while the return arrives later, so measured productivity can dip first.

Management is the dividing line

Among UK firms, those rated as well managed in 2020 were likelier to have adopted AI three years on, whatever their sector, location, size or age. Management practice, plus workforce and company characteristics, explains about 80% of the difference between generative AI use by US workers and by European ones.

Against five comparison countries (India, Brazil, South Africa, Colombia and the US), Britain has the biggest proportion of workers in roles where AI is more likely to support than replace them. The report links this partly to the size of its finance and professional services industries.

Looking Forward

The half-point to 0.7 estimate sits towards the lower end of the 0.4 to 1.3 point range the Treasury cited, from OECD work, in its 28 October Budget. In our view, the useful message for UK leaders is the report’s quickest win: deploy the AI tools already available in sectors where Britain is strong, including health and advanced manufacturing, rather than waiting for new ones. The 6% figure shows how few firms have got that far.