OpenAI just dropped its first-ever ‘state of enterprise AI’ report, and the numbers put hard data behind what many have suspected: the corporate AI experiment is rapidly becoming a corporate fixture. Drawing from anonymized usage data and a survey of 9,000 workers across nearly 100 companies, the report paints a picture of a workforce that isn’t just dabbling with chatbots but integrating them into daily operations at a startling pace.
Weekly messages in ChatGPT Enterprise have exploded roughly 8× over the past year, and the average worker is now sending 30% more messages than before. A more telling shift lies in a 19× year-to-date spike in the use of structured workflows like Projects and Custom GPTs. This signals a clear move away from casual, one-off queries toward repeatable, integrated business processes. The consumption of reasoning tokens—a proxy for using more intelligent models for complex work—has skyrocketed by about 320× in 12 months.
Workers aren’t just getting faster; they’re expanding their capabilities. The survey found that 75% of users say AI helps them complete tasks they previously couldn’t perform at all. Coding-related messages from non-technical staff jumped 36%, suggesting that AI is effectively lowering the barrier to entry for technical work. The personal time savings are significant, with employees reporting they claw back 40 to 60 minutes per day, and heavy users claiming more than 10 hours back each week.
A new divide is also emerging. ‘Frontier’ workers at the 95th percentile send 6× more messages than the median employee, and frontier firms log twice the messages per seat. The primary bottleneck, OpenAI argues, is no longer model capability—it’s organizational readiness to absorb the technology that now ships a new feature every three days.
