Enterprises see weak AI ROI
A KPMG-backed analysis covered by CIO finds many companies are adopting AI but still struggling to demonstrate measurable returns, creating a gap between purchase and proven value. (cio.com) The report says part of the problem is AI replacing services that were never properly measured, leaving buyers under pressure to justify tools on business terms. (cio.com)
Companies are pouring more money into artificial intelligence even as many still cannot show a clear return on that spending. (cio.com) KPMG’s Global Artificial Intelligence Pulse Survey found three out of four global leaders will prioritize artificial intelligence investment despite economic uncertainty, and its United States survey said organizations project average spending of $207 million over the next 12 months. (cio.com) (kpmg.com) The gap is not adoption alone. KPMG said 54% of organizations have integrated artificial intelligence agents into operations, but 65% of respondents said realizing return as those deployments scale is getting harder. (kpmg.com) Return on investment is a simple business test: did a tool save more money or make more money than it cost. CIO reported that many artificial intelligence projects are replacing services companies never measured well in the first place, which makes before-and-after comparisons hard to prove. (cio.com) That leaves technology chiefs in a bind. Boards are treating artificial intelligence as mandatory, but finance teams still want business cases stated in revenue, cost, productivity, or profit terms. (cio.com) (kpmg.com) The companies showing the strongest results are further along. KPMG said 82% of artificial intelligence leaders reported meaningful business value, versus 62% of peers that are still earlier in deployment. (cio.com) The survey points to several reasons returns stay fuzzy: skills gaps, trouble scaling pilots into routine work, and data privacy and cybersecurity risks. In the United Kingdom slice of the survey, 46% cited skills gaps and risk issues as the biggest barriers to demonstrating artificial intelligence-related return on investment. (ciodive.com) (kpmg.com) KPMG’s United Kingdom results also show where measurement breaks down. While 76% said they could measure productivity gains and 64% said they could measure profitability, only 14% said they were confident measuring returns from better analytics used by senior executives in decision-making. (kpmg.com) That helps explain why some executives are changing the standard. KPMG said 65% of United Kingdom respondents would keep investing in artificial intelligence regardless of tangible return on investment, framing the technology more as a long-term operating shift than a short-term software purchase. (kpmg.com) The pressure to show results has not gone away. KPMG’s 2025 United States pulse found 93% of leaders said generative artificial intelligence had improved competitive position, but its 2026 reporting shows the harder part is turning pilots and point wins into repeatable, enterprise-wide value. (kpmg.com 1) (kpmg.com 2) For large companies, the argument has shifted from whether to buy artificial intelligence to how to prove it changed the business. The spending is already here; the measurement system is still catching up. (cio.com) (kpmg.com)
