Trending Now - Risks of Seeking Health Advice From AI Chatbots
WHO:
OpenAI: Introduced ChatGPT Health in January 2026.
Anthropic: Offers similar features through its Claude chatbot.
Dr. Robert Wachter: Medical technology expert at the University of California, San Francisco.
Dr. Lloyd Minor: Dean of Stanford's medical school.
Ashwin Ramaswamy, MD: Lead author of a study at the Icahn School of Medicine at Mount Sinai.
Girish N. Nadkarni, MD: Chief AI officer of the Mount Sinai Health System.
Adam Mahdi: Lead author of a study at the Oxford Internet Institute.
WHAT:
ChatGPT Health is designed to analyze users' medical records, wellness apps, and wearable device data to answer health questions.
The program is currently on a waiting list for users.
A study published in the journal Nature Medicine on February 23, 2026, found that ChatGPT Health failed to recommend emergency care for a significant number of serious medical cases.
The study involved 960 interactions with ChatGPT Health across 60 clinical scenarios in 21 medical specialties.
ChatGPT Health reportedly has about 40 million users daily.
The study found that ChatGPT Health under-triaged over half of genuine emergencies and over-triaged roughly two-thirds of mild cases.
A 1,300-participant study by Oxford University found that AI chatbots did not lead to better health decisions compared to online searches or personal judgment.
WHERE:
Washington, D.C.: Location of the news articles.
University of California, San Francisco: Affiliation of Dr. Robert Wachter.
Stanford University: Affiliation of Dr. Lloyd Minor.
Icahn School of Medicine at Mount Sinai: Location of the study led by Ashwin Ramaswamy.
Oxford University: Location of the study led by Adam Mahdi.
WHEN:
January 2026: Launch of ChatGPT Health by OpenAI.
February 23, 2026: Publication date of the study in Nature Medicine.
2024: Year when the Oxford University study was conducted.
WHY:
Tech companies are offering health chatbots to meet the demand of hundreds of millions of people seeking health advice.
The aim of ChatGPT Health is to provide a more informed and confident navigation of health issues.
HOW:
ChatGPT Health uses large language models to summarize and explain complicated test results and analyze health trends.
The study at Mount Sinai involved creating clinical scenarios and assessing ChatGPT Health's responses based on urgency levels assigned by independent physicians.
Users must opt in to share their health information with AI companies, which is kept separate from other data and is subject to additional privacy protections.
OpenAI recently launched ChatGPT Health, a program designed to analyze users' medical records and wearable device data to answer health questions, although it is currently on a waiting list for users. A study published in Nature Medicine last week found that ChatGPT Health under-triaged over half of genuine emergencies while over-triaging about two-thirds of mild cases, raising concerns about its effectiveness in serious medical situations. Additionally, a separate study from Oxford University indicated that AI chatbots did not improve health decision-making compared to traditional online searches or personal judgment.