OpenAI releases data on 165 work activities people bring to ChatGPT

What the data covers

The release ties ChatGPT usage to 165 categories called Intermediate Work Activities (IWAs). OpenAI describes IWAs as commonly used categories of tasks for the kinds of work people do across occupations. They come from O*NET, a standardized framework that groups occupation-specific tasks into broader work activities shared across occupations.

The examples OpenAI gives range from preparing instructional materials and developing marketing content to programming and analyzing business data. The data is quarterly and runs from the first quarter of 2025 through the third quarter of 2026. An interactive chart lets users explore a specific country over time, moving from broad activity groups to more specific activities.

According to OpenAI, using a common framework helps researchers compare ChatGPT use with other types of occupational and employment data.

Countries are specializing

OpenAI says the data shows wide variation across countries in how people use ChatGPT for work. Work-related use also became less concentrated in the most common activities. In the company's reading, countries tended to specialize in their work over time, and different countries specialized in different types of work.

The figures: across 74 countries observed in every quarter, the ten most common activities made up an average of 51.1% of each country's classified work messages in Q3 2026, down from 56.0% in Q1 2025.

OpenAI notes that these differences might reflect variation in the kind of work done across countries or varying access to AI tools. It presents exploring those questions as one example of what researchers can pursue.

Comparing ChatGPT use with U.S. jobs

Because the data is linked to IWAs, researchers can compare ChatGPT messages with the activities performed across U.S. jobs. OpenAI provides paired visualizations that show the difference between how people use ChatGPT for work and the actual work done in the U.S. economy.

The two distributions differ substantially. OpenAI says this is largely because about 43% of the employment-weighted task mix for mapped U.S. occupations falls into Other IWAs, outside the individually reported categories. These include moving materials, operating vehicles, and maintaining health or medical records. According to the company, these categories did not meet its threshold for consistent representation in the original U.S. ChatGPT sample.

Scope and privacy, and what it means

The release builds on OpenAI Signals, which the company describes as its ongoing effort to share evidence about how people use AI. The data covers work-related messages on individual ChatGPT plans. It does not include enterprise messages or Codex use.

On privacy, OpenAI says researchers do not read individual conversations. Automated systems classify messages after identifying information is removed. The company publishes aggregated statistics, withholds results for small groups and adds statistical noise to protect individual messages.

For researchers and analysts, the release offers a standardized, country-level view of work-related AI use that can be matched to existing occupational data. OpenAI invites researchers to explore it, connect it with other sources, and investigate what these patterns mean for workers and businesses.