In a 69-page report published on August 11 examining enterprise adoption of ChatGPT, OpenAI presents a surface narrative of exponential AI usage growth across all seniority levels and job functions. The report highlights what it calls a "frontier gap," suggesting that companies using AI are pulling ahead of those that are not. However, a closer reading of the findings reveals a more nuanced—and less definitive—picture of AI's impact on business performance. The findings land at a time when enterprise buyers and analysts are under growing pressure to quantify returns on AI investments, which have consumed tens of billions in corporate spending over the past two years.

No Clear Link Between AI Usage and Revenue

On page 35 of the report, researchers acknowledge finding no statistically significant correlation between a company's revenue per employee and the intensity of AI usage among its workforce, whether measured by messages sent or tokens used. Revenue per employee is a standard benchmark used by analysts and investors to assess workforce productivity, making it a natural lens for evaluating whether AI tools are translating into measurable business outcomes.

"Revenue per employee is not meaningfully associated with output tokens per employee or messages per active user once other controls are included," the report states.

The researchers note that companies with higher revenue per employee tend to be early adopters of ChatGPT, and that companies using the technology more intensively tend to have higher revenue per employee in general. In other words, large, profitable companies are more likely to have adopted AI tools. What the study does not establish, however, is a causal relationship—namely, that greater AI usage directly produces greater revenue. The distinction between correlation and causation has been a persistent challenge in studies of workplace technology adoption, and this report does not resolve it.

Executives Use AI the Least

Another finding buried in the report is that senior employees are among the least intensive users of AI. According to data on page 29, executives send the fewest weekly messages per user, while early-career employees demonstrate by far the highest usage rates.

OpenAI CFO Sarah Friar emphasized this point in a LinkedIn post about the report: "For leaders, that's a reminder that competitive advantage comes from the people closest to the work. Listen to them, learn from them, and help the rest of the organization catch up."

Enterprise Usage Flatlined in Late 2025

The report also reveals that OpenAI's overall enterprise usage stagnated from approximately October 2025 through December 2025. A graph on page 26 depicting output token growth shows the line representing total growth nearly flat during that period. This stagnation coincided with Anthropic's Claude Code gaining significant traction in the corporate sector, becoming the preferred platform at many organizations.

Growth resumed sharply in January 2026, with the line turning upward into an exponential curve. "OpenAI's run rate in 2026 has been pretty incredible," one venture capitalist noted. OpenAI attributes the rebound to both new client acquisition and existing clients deepening their usage. CEO Sam Altman has been reorganizing the company around enterprise sales and discontinuing what the company referred to as "side quests", such as the video application Sora.

Leadership Shake-Up: New Chief Revenue Officer

On the same day the report's findings drew attention, OpenAI announced the appointment of Dali Rajic as its new Chief Revenue Officer, replacing Denise Dresser, who held the position for less than one year. Rajic's mandate will focus on accelerating customer adoption and helping businesses measure impact as the company advances toward its IPO. The leadership change signals that OpenAI views closing the measurement gap between usage and demonstrable business outcomes as central to its next phase of enterprise growth.

Report's Academic Contributors Were Paid by OpenAI

Of the report's five authors, two are academics and three are OpenAI employees. David Holtz, affiliated with Columbia Business School, and Prasanna Tambe, affiliated with Wharton at the University of Pennsylvania, are listed on the first page. However, a footnote clarifies that both "contributed to this work in their capacity as paid contractors for OpenAI."

The inclusion of academic researchers typically lends credibility and the appearance of independent institutional oversight, but the paid contractor arrangement complicates that perception. Disclosure of funding sources is standard practice in academic publishing, though the report itself was not published through a peer-reviewed journal.

Broader AI Industry Context

The report arrives amid a period of extraordinary capital flow in the AI sector. Two former OpenAI employees reportedly made approximately $10 million in a single day by selling shares through an internal tender offer, part of a broader buyback that Bloomberg reported totaled $7 billion across staff. Separately, Stockholm-based Lovable, founded just three years ago, recently doubled its valuation to $13.3 billion. Anthropic is also reportedly planning a $2 trillion IPO in October, which would be the largest in history.

Yet against this backdrop of surging valuations and capital deployment, the fundamental question of AI's return on investment for companies remains unresolved. OpenAI's own research, while optimistic in its framing, ultimately underscores that the evidence linking AI adoption directly to financial performance is not yet conclusive. As more AI companies approach public markets, the ability to demonstrate tangible enterprise ROI will likely face increasing scrutiny from public investors accustomed to revenue and margin metrics.

This story was originally featured on Fortune.com. The full OpenAI report is available here.