More AI use does not necessarily mean more revenue. New research from OpenAI found no meaningful statistical relationship between how heavily companies use ChatGPT and their revenue per employee, challenging the assumption that greater AI adoption automatically produces better financial results.
More AI Use Does Not Mean More Revenue
The 69-page working paper analyzed more than 17 million ChatGPT Enterprise messages from more than 1,500 organizations. Researchers compared AI usage with companies’ financial data while accounting for factors including company size and industry.
They found that revenue per employee was not meaningfully associated with either the number of messages employees sent or the amount of AI-generated tokens used per employee.
This is important, as companies increasingly try to justify large AI investments through productivity and financial returns. Usage alone does not show whether those investments are paying off.
Employees Are Using AI Across Their Jobs
AI use was spread across 60 categories of work, including writing, technical tasks, research, communication, sales, planning, legal work and data analysis.
Early-career employees were among the most intensive users, sending roughly eight to nine more messages per week than the average employee at their companies. Executives used the technology less frequently.
For employers, the findings point to a more difficult question than simply measuring adoption: Is AI actually improving the work that matters to the business?
The study does not find a financial payoff from AI usage alone. Companies will need to determine which applications produce measurable improvements in productivity, quality or business performance rather than assuming heavier usage will translate into higher revenue.
