A new National Bureau of Economic Research (NBER) working paper, How People Use ChatGPT (Chatterji, Cunningham, Deming, Hitzig, Ong, Shan, and Wadman, 2025, NBER Working Paper No. 34255), provides a detailed look at global ChatGPT usage. This post summarizes the authors’ findings and cites the source directly.

Key Findings

• Rapid Global Growth. By July 2025, ChatGPT reached more than 700 million weekly active users—about 10% of the global adult population. Adoption has accelerated in low- and middle-income countries.
• Work vs. Non-Work. Non-work messages grew from 53% in mid-2024 to over 70% in mid-2025. Work usage is most common among educated, highly paid professionals.
• Top Use Cases. Practical Guidance, Seeking Information, and Writing account for nearly 80% of conversations. At work, writing dominates, especially editing, summarizing, and translating user text.
• Programming is smaller than expected. Only about 4.2% of messages are programming-related.
• Interaction Types. Messages classify as Asking (~49%), Doing (~40%), and Expressing (~11%). At work, Doing is ~56% and is largely writing tasks.
• Work Activities. Most work-related use centers on (1) obtaining, documenting, and interpreting information, and (2) making decisions, solving problems, and thinking creatively.
• Demographics. The early male-skewed usage has closed; by mid-2025 usage is roughly balanced, slightly female-leaning. Nearly half of all messages come from users under 26. Higher education correlates with work use.
• User Satisfaction. Positive interactions outnumber negative ones by about 4:1. Asking messages are rated higher than Doing.

Why This Matters

The authors conclude that ChatGPT’s strongest economic value is as a decision-support tool. It helps people make choices, think through problems, and produce better writing. Unlike traditional search, ChatGPT generates tailored, actionable outputs that map to common work activities across jobs.

Full Reference

Chatterji, A., Cunningham, T., Deming, D. J., Hitzig, Z., Ong, C., Shan, C. Y., & Wadman, K. (2025). How People Use ChatGPT (NBER Working Paper No. 34255). National Bureau of Economic Research. https://www.nber.org/papers/w34255

Disclaimer

This post summarizes findings from the NBER working paper cited above. It does not represent original research by the author of this blog. All credit for the research and analysis belongs to the listed authors and to the National Bureau of Economic Research.

Categories: Artificial Intelligence

Tags: AI adoption, AI and work, AI demographics, AI research, chatgpt, Decision support, Generative AI, NBER, Productivity tools, Writing with AI