UK businesses lose GBP £67 billion on failed AI work

Large UK businesses are losing an estimated GBP £67 billion a year on transformation and AI initiatives that fail to deliver, according to research from Emergn. The consultancy estimated this equals an average 2.5% of annual revenue written off across large organisations.

Based on a survey of 700 senior leaders, the findings point to weak oversight rather than a lack of spending or ambition. Many organisations struggle to identify failing programmes early, find it difficult to stop them once under way, and cannot give boards a clear view of all live transformation and AI work.

Only 30% of UK leaders surveyed said stopping an underperforming programme is a normal part of how their organisation operates. A further 39% said action is taken only after significant time and money have already been committed, while 27% said they had seen programmes continue for no reason other than the amount already spent on them.

Board reporting also emerged as a weak point. Just 7% of UK leaders said every transformation and AI programme in their organisation is formally tracked and reported to the board, compared with 20% in the US. Emergn described this as the widest gap between the two markets in its research.

Most respondents also said they could not provide a complete live picture of current work. Only 32% said they could give their board a complete, real-time view of every transformation and AI programme on demand. The remaining 68% said they would need days, weeks or longer to assemble that information.

Governance gaps

The survey suggests companies are running large portfolios of change projects with limited senior-level visibility. UK organisations were found to be running an average of 6.4 transformation and AI initiatives at the same time, and 45% said they were running seven or more.

Some respondents reported minimal formal controls. About 11% said their organisation operates with no formal tracking or governance at all.

The problems appear before projects are fully approved. On average, only 16% of new ideas are formally rejected before substantial spending begins, suggesting most proposals move into delivery.

Leaders also described a gap between internal knowledge and what reaches the top of the organisation. A quarter of UK respondents said status reports present a more positive picture than the facts support, while 24% said bad news is softened as it moves up the chain.

Another 23% said staff stay quiet about programmes they believe are failing. A further 24% said evidence of failure is ignored or dismissed, a pattern UK leaders reported more often than their US counterparts.

AI scrutiny

The survey also points to reluctance around AI projects in particular. Nearly a quarter, or 23%, of UK leaders said senior leaders are reluctant to admit when an AI project has failed.

Emergn argued the issue is not primarily technical. Instead, it said the central problem lies in decision-making, governance and the willingness to stop initiatives that are not meeting expectations.

The GBP £67 billion estimate is presented as an indicative measure rather than an audited total. It is based on applying the reported 2.5% revenue loss across the combined annual turnover of large UK businesses, excluding the financial and insurance sector because comparable turnover data was not available.

That means the figure may understate the overall scale of losses if similar patterns hold in sectors not covered by the turnover calculation. Even so, the data suggests a substantial drag on investment at a time when many boards are increasing spending on AI and broader business change programmes.

Alex Adamopoulos, Chairman and Chief Executive Officer at Emergn, said: "Starting things is easy. Knowing when to stop them is the hard part, and it is where most of the money goes. Plenty of organisations can launch fifty initiatives. Very few can tell you, on any given Monday, which ones are paying off and which ones are quietly burning cash."

He said the issue reflects how companies govern investment choices rather than how many ideas they generate.

Adamopoulos added: "This is a decision-making problem, not an innovation one. Companies keep funding work they already know is failing, because of politics, because of what they have already spent, or because no one wants to be the person who admits it did not work. The winners in the AI era will not be the biggest spenders. They will be the ones with the discipline to act on the evidence, stop what is not working, and back what is. That discipline is fast becoming the real competitive advantage."