How People Use ChatGPT: Conversation-Level Evidence from India, Nigeria, Brazil, and Pakistan

Key Findings

• — 202,590 conversations from 1,252 users recruited via Clickworker; personal use accounts for 55-64% of conversations in every country, and coursework is about as common as work
• — Health and wellness is the largest unsupervised cluster in India (8.0%) and Brazil (9.2%); self-reflection and emotional conversations form a top-5 cluster in Brazil (5.5%)
• — The share of Expressing conversations rose from a few percent to roughly a fifth or more in every country over three years, while information-seeking declined only modestly
• — In a reading of 100 Expressing conversations per country, purely emotional conversations were rare: users typically disclosed personal context (health, relationships, academic pressure, money, religion) and then asked for advice or a task

Methodology Notes

Client-side scrubbing before upload; four classifiers including OpenAI's 24-category topic taxonomy and Asking/Doing/Expressing intent classifier, BERTopic with Gemini embeddings, and language detection. Sample is crowdworker-recruited, digitally literate and not representative (younger and more male than the adult population in India and Pakistan). No harm or safety outcome is measured. v1 submitted 2026-09-29 15:51 UTC (Thursday batch). Verified from arXiv abs and HTML render, both HTTP 200. Observation dates use month precision (exports span December 2022 to February 2026).

Sources