HMN 2025: How Many in US take into account AI-generated well being data helpful and dependable

Traditionally, people requested well being questions of their major well being care supplier. Confidence in that supplier as a supply of reliable well being data has been persistently excessive in recent times and was at 90% in April, in response to survey information from the Annenberg Public Policy Center (APPC) of the University of Pennsylvania.

The rise of the web created prepared entry to a second supply of skilled well being information—authorities well being authorities such because the Centers for Disease Control and Prevention (CDC), Food and Drug Administration (FDA), and National Institutes of Health (NIH). Although confidence in these sources eroded through the COVID-19 pandemic, relative to most different establishments confidence in these establishments has remained excessive, starting from 66%–68% in April for the three well being companies.

The creation of generative synthetic intelligence (AI) has added a 3rd supply of well being information and suggestions. Whether this supply is trusted—and, if that’s the case, to what extent—is an open query, however its presence is all however unavoidable. A Google seek for the reply to a well being query, for instance, will elicit a response led by an “AI Overview” of the search outcomes containing a disclaimer in small print on the finish: “This is for informational functions solely. For medical recommendation or prognosis, seek the advice of an expert. AI responses could embrace errors.”

Although Google notes that its AI summaries might not be correct, huge numbers of Americans are uncovered to those solutions and most take into account them dependable, in response to the most recent APPC well being survey. Conducted in April 2025 with over 1,600 U.S. adults, the survey finds that almost 8 in 10 adults say they’re probably to go surfing to reply a selected query about well being signs or a well being {condition}. Nearly two-thirds of this group say they’ve seen AI-generated responses. (Download the topline.)

Highlights

The Annenberg Public Policy Center’s April 2025 well being survey finds that:

• Most (79%) U.S. adults say they’re more likely to look on-line for the reply to a query a few well being symptom or {condition}.
• Three-quarters (75%) of people that search on-line say that AI-generated responses present them “typically” (45%) or “usually or extra” (31%) with the reply they want.
• Most Americans (63%) assume AI-generated well being data is considerably (55%) or very (8%) dependable.
• Nearly half (49%) are usually not comfy with well being care suppliers utilizing AI instruments quite than their {experience} alone when making selections about their care.

Almost 1 in 3 folks (31%) who use a web based search engine to seek out well being data say the AI-generated abstract response usually or at all times offers them the reply they want whereas one other 45% say it typically offers them the reply they want. Nearly two-thirds (63%) discover AI-generated outcomes to be considerably or very dependable. Yet on the similar time, nearly half of U.S. adults (49%) say they don’t seem to be comfy with well being care suppliers utilizing AI instruments quite than their {experience} alone when making selections about their care.

“Despite the disclaimers that accompany some AI-generated summaries, there may be potential for confusion and even hurt amongst weak people if they don’t seem to be conscious that these responses are usually not an alternative to the personalised skilled well being steerage that their well being care supplier can supply,” says Kathleen Hall Jamieson, director of the Annenberg Public Policy Center and co-editor of the guide “Realizing the Promise and Minimizing the Perils of AI for Science and the Scientific Community,” which is obtainable for gratis on-line.

APPC analysis analyst Laura A. Gibson added, “AI platforms are usually not essentially up to date in actual time and will comprise outdated data. Skepticism is warranted.”

Where do folks search well being data on-line?

While most individuals say they search on-line well being data, there is not a single supply or kind of supply to which they flip. The survey finds that the preferred sources are engines like google similar to Google or Bing, with 71% of respondents saying they usually or sometimes get well being data from these kinds of sources.

Following engines like google, the most typical sources folks report utilizing are on-line net sources (48%) like WebMD or healthline.com; web sites of huge medical analysis and scientific entities (44%) such because the Mayo Clinic and Cleveland Clinic; nonprofit organizations (37%) targeted on particular matters just like the American Cancer Society (cancer.org), American Heart Association (coronary heart.org), and American Diabetes Association (diabetes.org); and federal authorities sources (35%) such because the CDC, the NIH, and the FDA.

Fewer folks report getting data usually or sometimes from state or native authorities well being websites (14%), podcasts (18%), the World Health Organization (19%), or social media sources like Facebook, X (previously Twitter), TikTook, YouTube, or Reddit (24%).

Most say AI-generated well being data gives what they want and is dependable

Among customers who say they seek for well being data on-line, practically two-thirds (65%) report that they’ve seen AI-generated responses on the high of the outcomes on engines like google. On Google, for instance, that is labeled an “AI Overview,” and on Bing, as a “Copilot Answer.”

Seniors age 65 or older who say they seek for well being data on-line are considerably much less probably than youthful folks to say they’ve seen AI-generated well being responses atop their search outcomes. Under half (49%) of these age 65+ say they’ve seen AI well being responses, in contrast with 69% of 18- to 34-year-olds, 75% of 35- to 49-year-olds, and 63% of 50- to 64-year-olds.

Among those that have seen AI-generated responses on the high of outcomes when trying to find well being data, greater than 3 in 10 (31%) say AI at all times (4%) or usually (26%) gives them with “the reply you want.” Nearly half (45%) say AI typically offers them the response they want. About 1 in 5 (18%) say they hardly ever (13%) or by no means (6%) get the reply they want and 6% had been not sure.

Most individuals who seek for health-related data do not cease with an AI-generated response. About two-thirds (65%) say they at all times (31%) or usually (34%) proceed looking by following hyperlinks to particular web sites or different sources. Another quarter (24%) say they often comply with the hyperlinks, whereas 8% hardly ever or by no means transcend the AI-generated response.

Most individuals who seek for well being data on-line say they discover AI-generated well being data to be dependable. Nearly two-thirds (63%) discover the knowledge to be dependable, with 55% saying the knowledge is considerably dependable and eight% saying it is vitally dependable. 1 / 4 of on-line searchers (25%) say AI-generated well being data is usually unreliable.

Sharing on-line well being data with one’s well being care supplier

When visiting their physician or well being care supplier, few individuals who seek for well being data on-line (12%) say they at all times or usually speak with their physician or well being care supplier about well being data they noticed on-line. Two-fifths (40%) typically do and practically half say they achieve this hardly ever (29%) or by no means (17%).

Among individuals who commonly go to their well being care supplier and seek for on-line well being data, most agree (80%) that their well being care suppliers are typically good at explaining well being points and coverings to them. Nevertheless, respondents are blended about whether or not their well being care suppliers are dismissive or would not have the time to debate on-line well being data they create to their visits.

A 3rd (32%) say there’s often not sufficient time at medical appointments to speak with their suppliers about well being data they noticed on-line and 40% really feel there may be sufficient time, whereas 28% neither agree nor disagree. As for the suppliers’ receptivity to the net data introduced in by a affected person, 22% agree well being care suppliers will dismiss what they create in, however 37% disagree with that assertion. Four in 10 (41%) neither agree nor disagree.

Are you comfy with well being care suppliers utilizing AI instruments?

Nearly half of these surveyed (49%) say they don’t seem to be comfy with well being care suppliers utilizing well being care AI instruments quite than their {experience} alone when making selections in regards to the affected person’s care. More than a 3rd (36%), nonetheless, categorical no less than some consolation with their well being care suppliers utilizing AI instruments past their very own experience/{experience}.

Despite the widespread unease with well being care suppliers utilizing AI, greater than 4 in 10 folks surveyed (41%) assume their well being care supplier commonly makes use of well being care AI instruments typically (31%), usually (9%), or at all times (1%), quite than their {experience} alone when making selections about affected person care. Over 1 / 4 (26%) say their well being care suppliers hardly ever or by no means use AI, whereas a 3rd (33%) are usually not positive whether or not they use AI or not.

Many use sensible units to trace their well being information

Overall, 93% of respondents say they personal a smartphone or smartwatch. Adults 65 and older are considerably much less probably (83%) to personal one than youthful folks, who vary by age group from 98% to 95%.

About 6 in 10 (59%) sensible machine customers—cellphone or watch—report that they use an app on their machine that summarizes well being data, similar to steps per day, sleep time and sleep high quality, or coronary heart well being. Forty % of machine customers say they don’t use such an app on their machine. By age, these with a wise machine who’re between the ages of 18 and 34 (62%) and 35 and 49 (61%) are considerably extra more likely to say they use such an app than these age 65 or older (51%).

A majority of sensible machine customers (52%) say they’d be more likely to share a few of their abstract well being data with their physician or different well being care supplier. This contains the three% who say they achieve this already. More than a 3rd (36%) say they don’t seem to be more likely to share this data with their well being care supplier.

While a large variety of sensible machine customers don’t want to share their abstract well being care information with their docs or well being care suppliers, the overwhelming majority of survey respondents are comfy having their pharmacies ship them alerts about their prescriptions. Eight in 10 of respondents say they both already get alerts (31%) from their pharmacies or can be probably to join such alerts (49%).

The Annenberg science and public well being survey

The survey information come from the twenty fourth wave of a nationally consultant panel of 1,653 U.S. adults carried out for the Annenberg Public Policy Center by SSRS, an impartial market analysis firm. Most have been empaneled since April 2021. To account for attrition, replenishment samples have been added over time utilizing a random chance sampling design. The most up-to-date replenishment, in September 2024, added 360 respondents to the pattern. This wave of the Annenberg Science and Public Health (ASAPH) survey was fielded April 15–28, 2025. The margin of sampling error (MOE) is ± 3.4 proportion factors on the 95% confidence stage. All figures are rounded to the closest entire quantity and will not add to 100%. Combined subcategories could not add to totals within the topline and textual content resulting from rounding.