A large new survey found that most Americans want to hear about doctors using AI before it touches their care. Almost half of them have no idea whether it already happened at their last appointment.
The Gist
- 72% of US adults say it is extremely or very important to be told when AI is used in their care.
- 81% want to know if AI analyzed their medical scans or helped make a diagnosis.
- 46% have no idea whether AI has ever been used in their own care.
- 63% want more say in how AI gets used on them.
Have ChatGPT Recap This Article
ChatGPTWhat the survey actually asked 3,488 people
The Pew Research Center questioned 3,488 US adults between June 22 and June 28 this year, through its long-running opinion panel. The question underneath it all was simple. When a hospital or a clinic uses software that thinks, does the patient have a right to hear about it?
The answer came back clear. 72% called it extremely or very important to be told, 16% called it somewhat important, and only 7% said it did not really matter to them. That is a rare level of agreement for a survey about technology.
It also lands in a strange spot. Ordinary people are getting more comfortable typing health questions into a chatbot at home, and ChatGPT can already read the medical records you upload to it, yet the same people want a warning when the clinic does something similar without asking.
The gap makes sense once you look at who is in control. At home you decide what to paste and what to ignore. In an exam room, the software works on you, not for you, and nobody hands you a menu.
How much AI is already in the room
Here is the number that should stop you. 16% of respondents said a provider had used AI somewhere in their care, 37% said it had not, and 46% simply did not know. Nearly half the country cannot answer the question at all.
Even the people who knew about doctors using AI were not much clearer on the details. Among them, 22% said they understood how it was used extremely or very well, 44% said somewhat well, and 32% admitted they did not really follow what happened.
That fog feeds the wider mood. Public enthusiasm for this technology has been sliding for a while, and we covered how only 16% of Americans now describe themselves as positive about AI. A tool nobody explains rarely earns trust on its own.
If you want a plain walkthrough of how good these systems actually are at reading medical information, the AI Noobies team put a short video together on how chatbot answers stack up against a real doctor’s.
Keep learning on AI Noobies:
- Netflix GenRec Picks What You Watch Next
- The ChatGPT Apple Messages Plugin Sends Texts
- Context Window: Why ChatGPT Forgets What You Said
The difference between a decision and paperwork
People are not asking to be told about everything equally. The survey split AI uses into separate tasks, and the results form a neat ladder from the medical to the administrative.
At the top sit the things that shape what happens to your body. 81% want to be told if AI analyzed their scans, 81% if it helped make a diagnosis, and 80% if it explained their lab results.
Lower down, the demand softens. 72% want a heads up when AI takes the notes during an appointment, 64% when it handles a prescription refill, and 56% when it books the appointment itself.
Read as one line, that ladder says something useful. People care about medical judgment, not clerical work, and any clinic reading this survey now knows exactly which disclosure matters and which one is noise.
How to ask your own doctor this week
Wanting a say and having one are two different things. 53% of respondents felt they had not too much or no say over how AI is used in their care, while 33% felt they had at least some, and 63% said they want more input.
The practical move is a single question at your next appointment. Ask whether any part of your visit, from the notes to the scan reading, involved AI, and ask who checked the output afterwards.
That question feels ordinary once it is out. You already ask which lab ran a test or which specialist read a result, and doctors using AI is the same kind of question about who did the work.
Zooming out, this is the conversation heading your way over the coming months. Public trust in these tools was already shaky, as our piece on how few people trust AI chatbots for news showed, and doctors using AI is where that argument stops being abstract and starts being about your own body.
Stay tuned on AI Noobies.
About the Author
