Only 18% track AI ROI

Only 18% of professionals say their organizations track the return on investment from artificial intelligence tools, even as workplace use keeps spreading. (accountingtoday.com) The figure comes from Thomson Reuters Institute’s 2026 AI in Professional Services Report, based on more than 1,500 respondents across 27 countries in legal, tax, accounting, risk, fraud, and government work. The report says another 40% do not know whether their organization measures artificial intelligence return on investment at all. (thomsonreuters.com) Artificial intelligence use is moving faster than the accounting. Thomson Reuters said organization-wide use of generative artificial intelligence in professional services rose to 40% in 2026 from 22% in 2025. (thomsonreuters.com) The gap is not just whether companies measure returns, but what they count. Accounting Today reported that organizations that do track return on investment often focus on internal metrics such as employee use, not business measures such as revenue generation or client satisfaction. (accountingtoday.com) That pattern shows up outside professional services too. Gartner wrote in February 2026 that executives often rely on activity metrics such as productivity or adoption rates instead of financial outcomes such as cost reduction, revenue growth, or employee experience. (gartner.com) Big companies are still spending heavily while they sort out those measures. KPMG said in April 2025 that 130 United States business leaders at companies with at least $1 billion in annual revenue planned to invest nearly $114 million in generative artificial intelligence over the next 12 months, up from $89 million the prior quarter. (kpmg.com) KPMG also found that 97% of those leaders cited improved profitability as a return-on-investment metric and 94% cited productivity. In the same survey, 90% said investor pressure to demonstrate return on investment was important or very important, up from 68% in the prior quarter. (kpmg.com) Thomson Reuters said more than 80% of current generative artificial intelligence users engage with it weekly, and more than 90% expect it to become a central part of their workflow within five years. The report says those expectations are arriving before many firms have set clear success criteria. (thomsonreuters.com) The result is a workplace where adoption can be easy to count and business value can be harder to prove. For companies buying more artificial intelligence tools in 2026, the next test is whether usage dashboards turn into revenue, margin, or client results that finance teams can defend. (gartner.com)