Wharton study: companies can't prove their AI spending pays off

A University of Pennsylvania's Wharton School and GBK Collective study of 1,400 employees at firms with $50M+ revenue found that 40% of managers measure AI's impact in hours saved per employee and 37% in reduced errors, while only 27% track revenue from AI features. 44% expect AI to mainly boost productivity over the next one to two years, and just 8% expect new revenue or products.

• 40% of managers measure AI impact in hours saved per employee
• 37% track error reduction, only 27% track revenue from AI features
• 44% expect AI to mainly boost productivity in the next 1–2 years
• Only 8% expect AI to drive new revenue or products