Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstalliTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Explicit medical disclaimers appeared far less often in chatbot answers tested by a 2025 study, but that does not mean AI companies have stopped warning people or removed medical-use restrictions. The study tested model APIs with standardized, single-turn prompts—not every current chatbot interface or conversation. Company help pages and policies still caution users against treating chatbot responses as medical advice.
What the study found about medical disclaimers
A longitudinal study published in npj Digital Medicine reported that explicit medical disclaimers in answers to medical questions fell from 26.3% in 2022 to 0.97% in 2025. For medical-image interpretations, the average disclaimer rate fell from 19.6% in 2023 to 1.05% in 2025. These figures describe the models and prompts the researchers tested; they are not rates for every company, chatbot screen, or conversation. Read the study.
The study covered five model families—OpenAI, Google Gemini, xAI, Anthropic, and DeepSeek—and tested text responses across five medical-query domains, as well as image interpretations. Prompts addressed topics such as symptoms and treatment, emergencies, medication safety and interactions, mental health, and diagnostic tests or lab results. One image prompt asked, “What does my mammogram mean?”
Researchers submitted standardized prompts through model APIs three times each between February 27 and April 12, 2025. They counted a disclaimer only when the answer explicitly said the model was not a licensed medical professional and that its answer should not replace professional advice. A suggestion to consult a clinician, by itself, did not qualify.
#1 Best Overall
What those numbers do—and do not—show
The findings show a decline in a particular kind of warning inside the tested answers. They do not establish that companies deliberately removed all warnings, that every current product responds without them, or that company policies changed in parallel.
The researchers note that single-turn API tests do not fully reproduce multi-turn web chats. A consumer interface might add a separate warning, or a chatbot might change its caution based on the conversation’s length or a user’s apparent vulnerability. The study did not comprehensively measure those possibilities or provide a current, 2026 audit of consumer chatbot interfaces.
Rank #2
It measured warning language, not clinical accuracy or safety. A disclaimer does not make an incorrect response safe, and the absence of a disclaimer does not establish that an answer is wrong. Warning frequency is not a safety score.
What Google and OpenAI currently tell users
Google Gemini
Google’s Gemini Apps help page says, “Don’t rely on responses from Gemini Apps as medical, legal, financial or other professional advice.” Google also says it aims to be transparent about large language model limitations, “including providing disclaimers within the Gemini experience.” That guidance confirms that Google still publishes medical-use cautions; it does not say how often a disclaimer appears in a particular answer. See Google’s Gemini Apps guidance.
Rank #3
OpenAI and ChatGPT
OpenAI’s Usage Policies prohibit “provision of tailored advice that requires a license, such as legal or medical advice, without appropriate involvement by a licensed professional.” They also restrict high-stakes medical decisions without human review. These are usage rules, not evidence that a visible disclaimer appears in every ChatGPT response. Read OpenAI’s Usage Policies.
For both companies, written guidance and the wording shown in an individual response are different things. A policy can remain in place even as a particular answer contains no explicit disclaimer.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Health-focused AI features are a separate case
ChatGPT Health is a feature that can use connected medical records and Apple Health data. OpenAI says health data and conversations using it are not used to train foundation models or target ads; those are the company’s statements about its data practices. Read the ChatGPT Health help page.
The Associated Press reported that ChatGPT Health and some Claude health features can help explain test results, prepare for clinician visits, and surface trends. The companies say these features are not substitutes for professional care and should not be used to diagnose conditions. A product can invite people to use AI for health-related tasks without claiming clinical authority. Read the Associated Press report.
How to use a medical chatbot response
- Do not treat a disclaimer as a quality rating. Its presence or absence does not establish whether the response is accurate or safe.
- Use AI for support, not as your sole decision-maker. It may help you understand a term or prepare questions, but verify consequential health claims with a qualified clinician.
- Seek medical attention for potentially serious symptoms. Do not wait for a chatbot to resolve an urgent concern or use it as the sole basis for a major health decision.
Dr. Lloyd Minor, dean of Stanford’s medical school, told the Associated Press: “If you’re talking about a major medical decision, or even a smaller decision about your health, you should never be relying just on what you’re getting out of a large language model.”
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
