Cost visibility has been increasingly becoming both more complex and more critical to realising value.
In a survey by KPMG, amid increasing AI adoption scales and AI usage-based models, many organisations still lack a clear view of how spending accumulates.
The latest Global AI Pulse: Q2 2026 found that 23% of leaders struggle with usage-based costs, and 42% have only partial visibility into AI spending, compounded by challenges in understanding AI costs structures such as tokens, making it harder to predict, manage and optimise spending.
Organisations ought to put governance controls in place to better manage AI costs. According to the study, this includes monitoring and spending controls.
The research also revealed that 53% of leaders report having AI costs monitoring dashboards and embedding cost reviews as part of AI approval processes, enabling stronger control and decision making.
As a result, according to KPMG, these organisations are five times more likely to report established ROI (15 percent vs. 3 percent). These highlight that AI is no longer just a technical capability; it is a cost, margin and operating model priority.
While AI remains a top priority, as seen by 79% of leaders as a key investment area and with spending holding steady at $188M, there is greater emphasis being placed for global leaders to measure and demonstrate value as token economics add new complexity, driving a sharper focus on control, financial discipline and leadership accountability.
Further, AI adoption is accelerating, particularly in the driving-adoption phase, as it increased to 22% from 13% in the prior quarter, marking the largest shift on the AI maturity curve, signaling momentum.
However, KPMG said turning that momentum into measurable returns remains concentrated to only 7% of leaders who report establishing ROI, even as 24% face pressure to prove value to investors.
