One in seven Americans report trusting AI chatbots as much as they trust a doctor for health advice, according to a new survey by Talker Research for MD Live by Evernorth. The study, which surveyed 1,000 millennial parents and 1,000 general population Americans, explored how individuals are utilizing AI for their health and their children's health.
Among those who use AI for health purposes, the tendency to trust AI advice rivals that of medical professionals. The survey found that 60% of AI users would believe health advice generated by AI even if it conflicted with guidance from a human healthcare provider. Nearly two-thirds (63%) of these users admit to following AI-generated health advice without first consulting a healthcare provider.
A significant factor influencing this trust appears to be the way AI delivers information. Half of respondents indicated that a clear or reassuring response from an AI about their health is likely to be accepted without further questioning. These individuals frequently turn to AI for health queries, asking an average of nine health-related questions per week.
The primary reasons cited for relying on AI include its 24/7 availability, the speed of responses compared to seeing a doctor, and the reassurance it provides for health concerns. However, a considerable number of respondents, 28%, reported verifying AI-generated answers less than half the time.
For millennial parents using AI for health-related matters, the reliance is even more pronounced. Over half (58%) have acted on AI-generated health advice for their children without prior confirmation from a healthcare provider. Many expressed comfort with using AI for various decisions concerning their children's health, including evaluating minor symptoms (80%), determining the necessary level of care (71%), and seeking medication guidance (71%).
Despite this confidence, a large majority of respondents, 78%, failed to recognize that the wording of a question can influence the AI's response. More than half (53%) admitted to rephrasing their health questions when dissatisfied with the initial answer, with 27% stating they consistently do so. Many users also shape their questions to elicit specific or desired responses.
When seeking health information, users often omit crucial personal details. More than three-quarters of regular AI users (76%) do not provide their complete medical history or demographic information due to privacy concerns. Forty-two percent admitted to providing false information about the subject of the health question for privacy reasons.
This incomplete information sharing leads to a disconnect, with 59% of AI users later realizing the health information received was not entirely applicable to their situation. Nearly half (49%) proceeded to act on this information before realizing its limitations.
Concerns about receiving incorrect advice (34%) and the lack of human oversight (31%) were noted by respondents. Looking ahead, one in four individuals plan to increase their use of AI for healthcare decisions in the coming year. Additionally, 27% stated they would trust AI-generated health information more if it had been reviewed by a doctor.
Dr. Maggie Williams, medical director of virtual primary care at MD Live by Evernorth, commented on the survey's findings, suggesting that while increased engagement with health questions is positive, the process differs significantly from a doctor's approach. Physicians gather context, ask follow-up questions, and rule out possibilities before reaching a conclusion. The survey also highlighted a misunderstanding of AI's capabilities, with 22% believing AI exclusively uses trusted medical sources and an equal percentage thinking AI considers personal medical history even if not provided.
"As we consider a patient’s full care journey, it’s important to understand how AI and physicians can come together to provide the clearest picture to patients while ensuring appropriate safeguards and clinical oversight," Williams stated.