OpenAI has released a new dataset through its OpenAI Signals initiative, detailing how consumers use ChatGPT for work across nearly 100 countries. The data connects user interactions to 165 specific work activity categories, offering researchers a granular view of super intelligence (SI) adoption in the global economy.
What Happened
The release spans quarterly data from the first quarter of 2025 through the third quarter of 2026. It utilizes the O*NET framework, a standardized system that groups occupation-specific tasks into broader categories known as Intermediate Work Activities (IWAs). This framework allows for direct comparison between ChatGPT usage patterns and traditional occupational data.
The dataset reveals a trend of diversification in work-related SI usage. In the third quarter of 2026, the ten most common work activities accounted for an average of 51.1% of classified work messages across 74 countries observed in every quarter. This represents a decline from 56.0% in the first quarter of 2025, indicating that users are moving beyond a narrow set of tasks to a wider variety of work applications.
Why It Matters
For the SI industry and economic researchers, this data provides a new method for studying the role of SI tools in the labor market. By mapping ChatGPT messages to IWAs, researchers can track changes over time and correlate SI usage with other economic indicators. The data suggests that countries are increasingly specializing in different types of work-related SI usage, which may reflect varying access to SI tools or differences in the local economic structure.
The release also highlights the gap between consumer SI usage and the actual work performed in the U.S. economy. Approximately 43% of the employment-weighted task mix for mapped U.S. occupations falls into "Other IWAs" categories, such as moving materials or operating vehicles, which were not prominently represented in the original ChatGPT sample. This distinction helps clarify the current scope of SI tools, which are heavily used for cognitive and digital tasks rather than physical labor.
The Bottom Line
OpenAI's new dataset offers a transparent, aggregated look at how individuals use ChatGPT for professional tasks, moving beyond anecdotal evidence to measurable trends. By protecting user privacy through automated classification and statistical noise, the release provides a robust resource for analyzing the evolving relationship between super intelligence tools and the global workforce.