Americans Rely on AI for Health Queries Amid Accuracy Concerns

In a rapidly evolving digital landscape, a significant portion of the American populace is increasingly turning to artificial intelligence (AI) for health-related inquiries, despite ongoing concerns regarding the accuracy of information provided by these systems. According to a recent survey conducted by the Annenberg Public Policy Center (APPC) of the University of Pennsylvania, approximately 79% of U.S. adults reported they are likely to consult online sources, primarily search engines, for answers to health questions. This trend underscores a notable shift in how individuals seek health information, diverging from traditional reliance on healthcare providers.

The APPC survey, conducted in April 2025 with a sample of 1,653 U.S. adults, reveals that nearly two-thirds of respondents have encountered AI-generated responses during their online searches. Despite disclaimers indicating that such AI summaries are for informational purposes only, the findings illustrate a paradox where many individuals consider AI-generated health information to be reliable. Approximately 63% of respondents rated AI-generated health content as either somewhat or very reliable, with 45% stating that AI often or always provides the answers they seek.

Historically, confidence in personal healthcare providers has remained relatively high, with a reported trust level of 90% in April 2025, as indicated by APPC data. This trust extended to established health authorities such as the Centers for Disease Control and Prevention (CDC), the Food and Drug Administration (FDA), and the National Institutes of Health (NIH), which maintained a confidence range of 66% to 68% during the same period. However, the emergence of AI as a health information source introduces new complexities into the decision-making process for both patients and healthcare providers.

Kathleen Hall Jamieson, director of the APPC and co-editor of "Realizing the Promise and Minimizing the Perils of AI for Science and the Scientific Community" (Penn Press, 2024), warns that despite the disclaimers accompanying AI responses, the potential for confusion and harm exists, particularly among vulnerable populations. Jamieson emphasizes that AI responses should not replace the personalized guidance that healthcare professionals offer.

The survey also highlighted the diverse sources from which Americans acquire health information. While a significant majority, 71%, rely on search engines like Google or Bing, other popular resources include online health websites such as WebMD (48%), and federal resources like the CDC (35%). However, trust in these sources has fluctuated, particularly during the COVID-19 pandemic, which saw a decline in confidence in established health institutions.

Moreover, the generational divide is evident in the survey results, with younger individuals more likely to encounter and trust AI-generated health responses compared to seniors aged 65 and older. While 69% of individuals aged 18-34 reported seeing AI-generated health responses, only 49% of seniors reported the same. This disparity could reflect broader trends in technology adoption and digital literacy across age groups.

Despite the growing reliance on AI, many Americans express discomfort with their healthcare providers using AI tools in clinical decision-making. Nearly half (49%) of respondents indicated they are not comfortable with healthcare providers relying on AI outputs rather than their clinical experience. This skepticism is further compounded by the fact that only 12% of those who search for health information online frequently discuss their findings with their healthcare providers, suggesting a potential communication gap that could impact patient care.

As AI continues to integrate into healthcare, experts caution about the implications of unverified information on patient health decisions. Laura A. Gibson, a research analyst at APPC, notes that AI platforms may not always provide real-time updates, making skepticism about their reliability warranted.

In conclusion, the increasing reliance on AI for health inquiries represents a significant shift in patient behavior, highlighting the need for careful consideration of the accuracy and reliability of AI-generated information. As healthcare providers navigate this new landscape, fostering open communication about AI's role and limitations will be crucial in maintaining patient trust and ensuring informed decision-making. The future remains uncertain as both technology and health information dissemination evolve, but awareness and education will be paramount in bridging the gap between AI-generated insights and personalized patient care.