Almost two-thirds of Americans use artificial intelligence-based tools to seek medical information, but about one-third don’t trust them.
That’s according to a new online survey of 2,057 American adults conducted by Harris Poll on behalf of Merck Manual. It found that more than three in five Americans (62%) have used AI tools for medical information, such as medical chatbots such as ChatGPT, Google Assistant, Siri, and Google Gemini.
However, according to the survey, 90% of respondents said they are taking steps to verify the validity of information provided by AI.
Meanwhile, 32% said they do not trust medical information from AI.
Digging into the data, we found that the most common uses for these tools included getting information about a specific symptom or disease (33%) and asking questions about drug side effects or dosage (26%).
When asking for specific medical information, more than half (54%) of respondents said they would enter a list of symptoms to receive a diagnosis, and 44% would also include additional details such as vital signs, medical history, and other personal health data.
Use of AI for health information was highest among Gen Z and Millennials, with 77% of each group reporting using AI, followed by Gen X at 58% and Baby Boomers at 40%. Parents of children under 18 were also more likely to use these tools, with 83% reporting use compared to 52% of non-parents.
“Artificial intelligence is a powerful tool for accessing and organizing information across all types of tasks, and health information retrieval is no exception,” Merck Manual Editor-in-Chief Sandy Falk, MD, said in an April 23 release.
“The challenge is to ensure that the information is accurate, reliable and provided in the right context.”
The first Merck Manual was published by Merck & Co. in 1899 as a pocket-sized reference work for physicians and pharmacists. It is now the leading online resource for medical professionals and consumers seeking medical information.
