UK businesses waste £67bn a year on failed AI projects

TL;DR:

• Large UK businesses lose an estimated £67bn a year on transformation and AI work that fails to deliver, according to research from consultancy Emergn — around 2.5% of annual revenue.
• The survey of 700 senior leaders points to weak oversight: only 7% say every AI programme is formally tracked and reported to the board, against 20% in the US.
• Just 30% treat stopping an underperforming programme as normal practice; 23% say leaders are reluctant to admit an AI project has failed.

The problem with UK AI spending, new research suggests, is not the amount but the discipline behind it. Emergn estimates that large organisations write off roughly £67bn a year on change and AI initiatives that miss their goals, based on applying a reported 2.5% revenue loss across big UK firms outside financial services.

A governance problem, not a technology one

The figures describe boards flying blind. Only 32% of leaders said they could give a complete, real-time view of every live AI and transformation programme on demand; the rest would need days or weeks to assemble it. UK firms run an average of 6.4 such initiatives at once, and 11% admit to no formal tracking or governance at all. Sunk-cost thinking runs deep — 27% have seen programmes continue solely because of what had already been spent.

That candour gap compounds the waste. A quarter of leaders said status reports paint a rosier picture than the facts support, and 23% said staff stay quiet about programmes they believe are failing — a pattern reported more often in the UK than the US.

The findings sharpen a theme running through recent UK data: enthusiasm outpacing results. Separate research this month found most UK firms use AI while calling it irrelevant, and that adoption gains cluster among a narrow group of workers. Emergn’s £67bn is an indicative estimate, not an audited total, and may understate losses in the sectors it excludes.

Looking forward

Emergn chief executive Alex Adamopoulos argued the winners “will not be the biggest spenders” but the firms with the discipline to stop what is not working. For UK boards under pressure to show AI returns, the message is that better project governance — clear baselines, honest reporting and the will to cancel — may matter more than the next tool.