In a recent study published in the journal Nature Health, researchers at Microsoft AI examined over 500,000 de-identified health-related conversations with Microsoft Copilot, revealing a growing trend of individuals using AI chatbots for inquiries that extend beyond basic health facts.
Health-related topics have emerged as a significant area of engagement for users of artificial intelligence (AI) chatbots, particularly those powered by large language models (LLMs) like ChatGPT and Copilot. These systems are increasingly utilized by individuals seeking assistance with a range of issues, from initial symptom assessment to queries about medications and interactions with healthcare providers. This evolution signifies a shift in the way people interact with digital technology and access health information.
The study focused on conversations occurring in January 2026, analyzing a random sample drawn from daily interactions with Copilot. Each conversation was categorized by general topic and intent, with those related to “health and fitness” included in the dataset. Researchers utilized an LLM-based classifier to assign conversations to one of twelve health intent categories, subsequently applying a clustering method to a sub-sample of 10,000 conversations for deeper analysis.
Findings indicated that the analytic dataset comprised 617,827 conversations classified under health and fitness. Approximately 41% of these discussions were categorized under health information and education, which included general queries about nutrition, medical conditions, and medication functions. Notably, even general inquiries may reflect personal health concerns, suggesting that the true proportion of personal queries could be higher than reported.
Additionally, a significant number of queries targeted specific conditions and treatments rather than general health information, which implies that users often seek general knowledge to inform personal health decisions. Use patterns varied markedly based on the device employed; for instance, mobile interactions peaked during nighttime, while desktop usage was more prevalent throughout the day. This divergence points to differing user engagement strategies depending on the platform.
Excluding the health information and education category, which accounted for around 40% of conversations across both device types, the researchers noted that personal and professional intents appeared in distinct patterns. Queries related to academic support and research represented 16.9% of desktop interactions but only 5.3% on mobile. Conversely, questions pertaining to symptoms and health concerns were more frequent on mobile (15.9%) than desktop (6.9%).
The analysis further revealed that discussions on personal health increased during evening hours, a trend consistent with prior research indicating a diurnal rhythm in emotional health, where negative feelings tend to peak at night. Personal health queries—including symptom descriptions and emotional well-being—comprised nearly one-fifth of interactions, while many users sought assistance navigating the healthcare system, such as finding providers and managing appointments.
The study’s limitations include its reliance on Copilot logs from a single month, which may not encompass seasonal variations, and the focus on inquiries rather than outcomes. As such, it remains unclear whether the information provided led to improved decision-making or subsequent care. Future research aims to explore the efficacy of AI-driven health information in aiding users.
As AI chatbots like Copilot become integral to health-related inquiries, their role in reducing administrative friction and enhancing patient empowerment in healthcare is becoming increasingly significant. The findings underscore the importance of understanding user intent within this evolving landscape, providing valuable insights for the development of more effective AI tools in healthcare.
Journal reference: Costa-Gomes, B., Tolmachev, P., Taysom, E., Sounderajah, V., Richardson, H., Schoenegger, P., Liu, X., Nour, M. M., Spielman, S., Way, S. F., Shah, Y., Bhaskar, M., Nori, H., Kelly, C., Hames, P., Gross, B., Suleyman, M., & King, D. (2026). Public use of a generalist LLM chatbot for health queries. Nature Health, 1-8. DOI: 10.1038/s44360-026-00117-x, https://www.nature.com/articles/s44360-026-00117-x
See also
AI Study Reveals Generated Faces Indistinguishable from Real Photos, Erodes Trust in Visual Media
Gen AI Revolutionizes Market Research, Transforming $140B Industry Dynamics
Researchers Unlock Light-Based AI Operations for Significant Energy Efficiency Gains
Tempus AI Reports $334M Earnings Surge, Unveils Lymphoma Research Partnership
Iaroslav Argunov Reveals Big Data Methodology Boosting Construction Profits by Billions
