Companies Pause 49% of AI Projects as Cost Controls Lift Measurable ROI to 15%
Updated
Updated · heise online · Jul 7
Companies Pause 49% of AI Projects as Cost Controls Lift Measurable ROI to 15%
3 articles · Updated · heise online · Jul 7
Summary
49% of companies have slowed, paused or scaled back AI projects over cost concerns, even as 79% of executives still rank AI a top area for new spending, KPMG found.
Consumption-based billing is driving the pullback: 42% said they have only a partial view of AI spending, about a third struggle to understand cost structures, and 23% cited problems with usage-based charges.
54% now build cost checks into AI project approvals and 53% use dashboards to track spending; firms with dedicated controls reported measurable AI ROI far more often—15% versus 3% without them.
22% are considering cheaper models, reflecting a broader shift from experimentation to tighter financial discipline as companies face pressure to prove AI value to investors.
As companies slash AI budgets, are they accidentally killing their most valuable and successful projects?
Are cheaper AI models the key to profit, or a gateway to even greater, uncontrolled spending?
What is 'shadow AI,' and how could your employees' unapproved tools be costing the company millions?
From Hype to Hard Numbers: The 2026 Crisis and Transformation in AI ROI
Overview
By July 2026, the initial excitement around AI investment has faded, replaced by a strong demand for real, measurable returns. The era of 'AI at Any Cost' is over, as organizations now expect clear value from AI projects. Many companies are seeing their AI initiatives stall and fail to deliver the returns they hoped for. Over the past two years, tolerance for experimental projects has dropped sharply. Businesses are no longer willing to fund AI based only on vague promises; instead, every project must show tangible results and align with specific goals.