Public use of a generalist LLM chatbot for health queries
Peer-reviewed analysis of over 500,000 de-identified health-related Microsoft Copilot consumer conversations from January 2026 (N=617,827 after exclusions), classified with a 12-category hierarchical intent taxonomy using privacy-preserving LLM-based classification validated against expert annotation. Characterizes what people ask a generalist AI assistant about health: personal symptom assessment, condition information, caregiving queries on behalf of others, and emotional-wellbeing conversations that concentrate in evening and night hours when traditional healthcare access is most limited.
Key Findings
- The emotional-wellbeing share of conversations rises by more than half from 3.3% in the morning (06:00-12:00) to 5.2% overnight (00:00-06:00), the hours when professional care is least available
- One in seven personal-health conversations is on behalf of someone else (a child, aging parent, or partner) — 14.5% of symptom questions concern a dependent — positioning conversational AI as a caregiving tool
- LLM-based intent classification agreed with human expert annotators in 84% of a 131-conversation review sample; mobile usage concentrates on personal health concerns while desktop is dominated by research/professional use
Methodology Notes
Nature Health, published online 2026-04-16, DOI 10.1038/s44360-026-00117-x, open access (CC BY 4.0). All 18 authors are affiliated with Microsoft AI (Redmond) — a vendor analysis of its own product's de-identified consumer data. Peer-reviewed version of the Microsoft report "How people use Copilot for Health" (microsoft.com/en-us/research, early 2026). Verified by fetching the open-access article page directly; venue/date/authors cross-checked against Crossref.
Sources
Nature Health article page (open access) (primary)
Microsoft Research report page (earlier report form)
Archived snapshot (Wayback Machine) — preserved against link rot
Authors
Beatriz Costa-Gomes, Pavel Tolmachev, Eloise Taysom, Viknesh Sounderajah, Hannah Richardson, Philipp Schoenegger
Tags
Cite This
APA
Beatriz Costa-Gomes et al. (2026). Public use of a generalist LLM chatbot for health queries. Nature Health. https://www.nature.com/articles/s44360-026-00117-x
Related Insights
Investigating Affective Use and Emotional Well-being on ChatGPT
OpenAI; MIT Media Lab · 4 Apr 2025
How people use Claude for support, advice, and companionship
Anthropic · 27 Jun 2025
How people ask Claude for personal guidance
Anthropic · 30 Apr 2026
Real-world use of large language models for mental health in 2024
npj Digital Medicine · 1 Aug 2026