OpenAI report finds no link between AI use and revenue gains

On August 11, OpenAI released a 69-page report analyzing enterprise adoption of ChatGPT, framing it as evidence that companies using AI are pulling ahead of those who do not. The headline story is one of exponential usage growth across all job levels and functions, and a 'frontier gap' favoring early AI adopters. Yet buried on page 35 is a finding that threatens to undermine that narrative: the researchers report no statistically significant correlation between revenue per employee and how much employees use AI, as measured by messages sent and tokens used. The report acknowledges that companies with already-higher revenue per employee tend to be early ChatGPT adopters, and that heavy AI users tend to work at higher-revenue firms in general. But it does not establish that increased AI use causes increased revenue.

The report also exposes an adoption skew. Data on page 29 show that executives use AI the least of all employee groups, posting the lowest weekly messages per user. Early-career employees, by contrast, have by far the most usage. OpenAI CFO Sarah Friar highlighted this finding on LinkedIn, framing it as evidence that competitive advantage comes from frontline workers, and urging leaders to listen to them and help the rest of the organization catch up.

A closer look at OpenAI's own business performance during the report period reveals further complications. A graph on page 26 depicting total output token growth shows the line nearly flat from October 2025 through December 2025—a three-month stall during which Anthropic's Claude Code became the go-to platform at many enterprises. The chart then thrusts upward into an exponential curve in January 2026, but ends by March 2026, leaving the trajectory beyond that point unclear.

In response, OpenAI has made an aggressive executive move. On the same day it published the report, the company announced it hired Dali Rajic as Chief Revenue Officer, replacing Denise Dresser after less than one year in the role. Rajic's mandate is to accelerate customer adoption and help businesses measure impact as OpenAI prepares for a potential IPO. CEO Sam Altman has also been reorganizing the company around enterprise sales, winding down what OpenAI called 'side quests,' such as the video app Sora.

The report carries one additional caveat: two of the five authors are academics—David Holtz from Columbia Business School and Prasanna Tambe from Wharton—who contributed as paid contractors for OpenAI. The other three are OpenAI employees. While including outside academics typically signals credibility and research independence, the fact that both external researchers were paid by the company muddies that implication. The article notes it remains unclear whether the researchers uncovered additional findings they chose not to publish, a standard practice in academic work but one rendered opaque here by the contractual relationship.