OpenAI Releases Data on ChatGPT Use Across 165 Work Activities

OpenAI has released new OpenAI Signals data examining how consumers use ChatGPT for work across nearly 100 countries. The dataset maps ChatGPT work-related messages to 165 Intermediate Work Activities, a standardized set of task categories derived from O*NET that span occupations and sectors. The quarterly series runs from the first quarter of 2025 through the third quarter of 2026, giving researchers a basis to compare usage patterns across countries, track trends over time, and link AI adoption to economic indicators. The release is intended for external study rather than product benchmarking. The data covers work-related messages from individual ChatGPT plans and excludes enterprise messages and Codex use. To protect privacy, OpenAI said automated systems classify messages after removing identifying information. The published data includes only aggregated statistics, withholds results for small groups, and adds statistical noise. The release reveals wide variation in how people use ChatGPT for work. Over time, work-related use became less concentrated in the most common activities, with countries appearing to specialize in different kinds of work. These patterns may reflect differences in labor markets, occupational mix, or access to AI tools. When compared with employment-weighted tasks mapped to U.S. jobs, ChatGPT use differed substantially. A major reason is that about 43 percent of the employment-weighted task mix for mapped U.S. occupations falls into broader Other Intermediate Work Activities, such as moving materials, operating vehicles, and maintaining health or medical records. Those categories were not consistently represented in the original U.S. ChatGPT sample.