KPMG finds firms rushing into AI but unable to prove value

TL;DR:

  • KPMG’s Global AI Pulse for Q2 2026 finds 22% of organisations now embed AI in daily workflows, up nine points in a quarter.
  • Yet almost half of businesses have paused or scaled back AI projects after deciding the cost outweighed the value.
  • Firms with full visibility of AI running costs report established returns at five times the rate of those without.

Adoption is accelerating but proof of value is not, according to KPMG’s latest Global AI Pulse. The consultancy found 22% of organisations have reached the “driving adoption” stage, embedding AI in everyday work — a nine-point jump on the previous quarter — even as cost pressure forces many to retreat.

The measurement gap

The report’s sharpest finding is financial: nearly half of businesses have paused or trimmed AI initiatives after concluding the spend was not justified, while those with clear sight of operating costs are five times more likely to show established returns. Specialists in finance, payments and cyber security told IT Brief the common failure is deploying fast without the discipline to prove worth. “More firms are implementing AI into everyday workflows, but that does not mean they are creating value,” said Oliver St Clair-Stannard of RedCompass Labs, arguing that regulated sectors must be able to show what AI produced, who signed it off and whether it would survive an audit.

Expense-software firm Rydoo made the same point about finance leadership: “confidence doesn’t equal proof”, and most businesses cannot say with precision what their AI actually costs to run. The findings echo a growing UK theme that trust and verification, not raw capability, now gate AI at scale, reinforcing earlier research naming trust as the real barrier.

Looking forward

For UK businesses, the message is that cost visibility is becoming the dividing line between AI as an investment and AI as an unmanaged operating expense. Finance functions are being pulled into decisions once left to technology teams, and boards are being urged to apply the same rigour to AI spend as to any strategic project. As deployments move deeper into critical systems, auditability and cost tracking look set to matter as much as the models themselves.

Share this article