OpenAI: Enterprise ChatGPT Messages Up 8x as AI Use Deepens

Weekly messages in ChatGPT Enterprise grew roughly 8x over the past year, OpenAI reports in its first "State of Enterprise AI" study, published today. The figure anchors a report that combines deidentified, aggregated usage data from OpenAI's enterprise customers with a survey of 9,000 workers across almost 100 enterprises.

The stakes are straightforward. ChatGPT now serves more than 800 million weekly users, and OpenAI argues that this consumer adoption has created "a powerful flywheel, accelerating the pace at which AI is being brought into work and professional settings." The company frames the moment with a historical analogy: general purpose technologies, from steam engines to semiconductors, create significant economic value only after firms translate underlying capabilities into scaled use cases. "Enterprise AI now appears to be entering this phase," the report states.

Adoption is deepening, not just spreading

The headline numbers describe intensification as much as breadth. Beyond the 8x growth in weekly enterprise messages, the average worker sends 30% more messages than a year ago.

Usage of structured workflows — Projects and Custom GPTs — has increased 19x year-to-date. OpenAI reads this as a shift from casual querying to integrated, repeatable processes. The most striking figure concerns reasoning: average reasoning token consumption per organization has grown approximately 320x in the past 12 months, which the company attributes to more intelligent models being systematically integrated into expanding products and services.

OpenAI describes growth on two margins. The extensive margin means more workers adopting AI. The intensive margin means existing users going deeper. Both are expanding simultaneously.

Technology, healthcare and manufacturing lead

Every sector shows traction, but the fastest-growing are technology, healthcare, and manufacturing. Professional services, finance, and technology operate at the largest scale.

Geographically, the fastest-growing business customer bases are Australia, Brazil, the Netherlands, and France, each exceeding 140% year-over-year growth. International API customer growth has surpassed 70% over the last six months, and Japan hosts the largest number of corporate API customers outside the United States.

Workers report time savings of 40-60 minutes a day

The survey findings point to measurable self-reported value. Three-quarters of workers say using AI at work has improved either the speed or quality of their output. Respondents report saving 40-60 minutes per day, and heavy users report more than 10 hours per week.

Department-level results from the survey:

• 87% of IT workers report faster IT issue resolution.
• 85% of marketing and product users report faster campaign execution.
• 75% of HR professionals report improved employee engagement.
• 73% of engineers report faster code delivery.

OpenAI's interpretation goes beyond speed. "Crucially, AI is not just helping people do the same work faster—it is enabling people to do new kinds of work," the report says. Coding-related messages increased 36% among workers outside technical functions. And 75% of users say they can now complete tasks they previously could not perform. The report's framing: "AI reduces the gap between intent and execution," letting individuals translate ideas into concrete outputs regardless of specialization or technical expertise.

A widening gap between frontier users and the median

The distribution of usage is highly uneven, and OpenAI says the gap is widening. Frontier workers at the 95th percentile send 6x more messages than the median employee and engage more intensively with advanced capabilities. Frontier firms send 2x more messages per seat and show deeper integration of AI across teams.

These patterns compound, according to the report, because time savings increase as users consume more intelligence and engage across more distinct tasks. Heavy users get more back per hour invested.

The pace of change adds pressure. OpenAI says it releases a new feature or capability roughly every three days, and the report argues that the primary constraints for organizations are "no longer model performance or tooling, but rather organizational readiness and implementation."

A benchmark, and a sales document

OpenAI positions the report as more than description: it benchmarks how leading enterprises deploy AI today, where they realize value, and how deeper integration compounds impact over time. The company encourages executives, operators, and builders to engage with the findings and invites feedback on open questions.

One caveat for readers weighing the numbers: the data comes from OpenAI's own customers and an OpenAI-conducted survey. Self-reported productivity gains, drawn from the population of firms already paying for enterprise AI, may not generalize. The company says all data points were deidentified and aggregated to preserve privacy.

OpenAI plans to keep publishing as the field shifts. The report's closing argument — that organizational readiness, not model capability, now gates adoption — sets up the next phase of enterprise AI competition, where vendors will increasingly sell implementation expertise alongside intelligence.