Executives Report AI Productivity Stagnation Due to Verification Tax

Business executives and researchers report a lack of significant productivity gains from artificial intelligence despite widespread adoption. This stagnation is driven by a verification tax, where the time saved through AI automation is consumed by the human effort required to check, correct, and rewrite low-quality output.

Data from a poll of 6,000 senior business executives co-published by the Federal Reserve Bank of Atlanta, Stanford University, and other institutions indicate that approximately 90 percent of senior executives see no impact on employment or productivity. A separate Workday survey of 3,200 employees and leaders found that staff spend an average of 37 percent of their AI-saved time on verification.

Experts note that while AI can rapidly generate content such as code or hypotheses, the human capacity to review and integrate that work remains a binding constraint. This shift moves bottlenecks further downstream in the production process. In response to these challenges, KPMG U.S. maintains human checks in every automated process to ensure quality.