Google AI Overviews cite YouTube more than any medical site for health queries, study suggests | Google | The Guardian

Skip to main contentSkip to navigationClose dialogue1/2Next imagePrevious imageToggle captionSkip to navigationPrint subscriptionsNewsletters Sign inUSUS editionUK editionAustralia editionEurope editionInternational editionThe Guardian - Back to homeThe GuardianNewsOpinionSportCultureLifestyleShow moreHide expanded menuNewsView all NewsUS newsUS politicsWorld newsClimate crisisMiddle EastUkraineUS immigrationSoccerBusinessEnvironmentTechScienceNewslettersWellnessOpinionView all OpinionThe Guardian viewColumnistsLettersOpinion videosCartoonsSportView all SportSoccerNFLTennisMLBMLSNBAWNBANHLF1GolfCultureView all CultureFilmBooksMusicArt & designTV & radioStageClassicalGamesLifestyleView all LifestyleThe FilterWellnessFashionFoodRecipesLove & sexHome & gardenHealth & fitnessFamilyTravelMoneySearch input google-search SearchSupport usPrint subscriptionsNewslettersDownload the appSearch jobsDigital ArchiveGuardian LicensingLive eventsAbout UsThe Guardian appVideoPodcastsPicturesInside the GuardianGuardian WeeklyCrosswordsWordiplyCorrectionsTipsSearch input google-search SearchSearch jobsDigital ArchiveGuardian LicensingLive eventsAbout UsUKUS politicsWorldClimate crisisMiddle EastUkraineFootballNewslettersBusinessEnvironmentUK politicsScienceTechGlobal developmentObituaries No hospital network, government health portal, medical association or academic institution came close to YouTube’s number of citations, the researchers said. Composite: Guardian Design/Getty ImagesView image in fullscreenNo hospital network, government health portal, medical association or academic institution came close to YouTube’s number of citations, the researchers said. Composite: Guardian Design/Getty ImagesGoogleGoogle AI Overviews cite YouTube more than any medical site for health queries, study suggestsExclusive: German research into responses to health queries raises fresh questions about summaries seen by 2bn people a month
How the ‘confident authority’ of AI Overviews is putting public health at riskAndrew Gregory Health editorSat 24 Jan 2026 12.00 ESTLast modified on Sun 25 Jan 2026 21.30 ESTShareGoogle’s search feature AI Overviews cites YouTube more than any medical website when answering queries about health conditions, according to research that raises fresh questions about a tool seen by 2 billion people each month.The company has said its AI summaries, which appear at the top of search results and use generative AI to answer questions from users, are “reliable” and cite reputable medical sources such as the Centers for Disease Control and Prevention and the Mayo Clinic.However, a study that analysed responses to more than 50,000 health queries, captured using Google searches from Berlin, found the top cited source was YouTube. The video-sharing platform is the world’s second most visited website, after Google itself, and is owned by Google.Researchers at SE Ranking, a search engine optimisation platform, found YouTube made up 4.43% of all AI Overview citations. No hospital network, government health portal, medical association or academic institution came close to that number, they said.“This matters because YouTube is not a medical publisher,” the researchers wrote. “It is a general-purpose video platform. Anyone can upload content there (eg board-certified physicians, hospital channels, but also wellness influencers, life coaches, and creators with no medical training at all).”Google told the Guardian that AI Overviews was designed to surface high-quality content from reputable sources, regardless of format, and a variety of credible health authorities and licensed medical professionals created content on YouTube. The study’s findings could not be extrapolated to other regions as it was conducted using German-language queries in Germany, it said.The research comes after a Guardian investigation found people were being put at risk of harm by false and misleading health information in Google AI Overviews responses.In one case that experts said was “dangerous” and “alarming”, Google provided bogus information about crucial liver function tests that could have left people with serious liver disease wrongly thinking they were healthy. The company later removed AI Overviews for some but not all medical searches.The SE Ranking study analysed 50,807 healthcare-related prompts and keywords to see which sources AI Overviews relied on when generating answers.They chose Germany because its healthcare system is strictly regulated by a mix of German and EU directives, standards and safety regulations. “If AI systems rely heavily on non-medical or non-authoritative sources even in such an environment, it suggests the issue may extend beyond any single country,” they wrote.AI Overviews surfaced on more than 82% of health searches, the researchers said. When they looked at which sources AI Overviews relied on most often for health-related answers, one result stood out immediately, they said. The single most cited domain was YouTube with 20,621 citations out of a total of 465,823.View image in fullscreenResearchers at SE Ranking found YouTube made up 4.43% of all AI Overview citations. Photograph: Adam Vaughan/EPAThe next most cited source was NDR.de, with 14,158 citations (3.04%). The German public broadcaster produces health-related content alongside news, documentaries and entertainment. In third place was a medical reference site, Msdmanuals.com with 9,711 citations (2.08%).The fourth most cited source was Germany’s largest consumer health portal, Netdoktor.de, with 7,519 citations (1.61%). The fifth most cited source was a career platform for doctors, Praktischarzt.de, with 7,145 citations (1.53%).The researchers acknowledged limitations to their study. It was conducted as a one-time snapshot in December 2025, using German-language queries that reflected how users in Germany typically search for health information.Results could vary over time, by region, and by the phrasing of questions. However, even with those caveats, the findings still prompted alarm.Hannah van Kolfschooten, a researcher specialising in AI, health and law at the University of Basel who was not involved with the research, said: “This study provides empirical evidence that the risks posed by AI Overviews for health are structural, not anecdotal. It becomes difficult for Google to argue that misleading or harmful health outputs are rare cases.“Instead, the findings show that these risks are embedded in the way AI Overviews are designed. In particular, the heavy reliance on YouTube rather than on public health authorities or medical institutions suggests that visibility and popularity, rather than medical reliability, is the central driver for health knowledge.”A Google spokesperson said: “The implication that AI Overviews provide unreliable information is refuted by the report’s own data, which shows that the most cited domains in AI Overviews are reputable websites. And from what we’ve seen in the published findings, AI Overviews cite expert YouTube content from hospitals and clinics.”Google said the study showed that of the 25 most cited YouTube videos, 96% were from medical channels. However, the researchers cautioned that these videos represented fewer than 1% of all the YouTube links cited by AI Overviews on health.“Most of them (24 out of 25) come from medical-related channels like hospitals, clinics and health organisations,” the researchers wrote. “On top of that, 21 of the 25 videos clearly note that the content was created by a licensed or trusted source.“So at first glance it looks pretty reassuring. But it’s important to remember that these 25 videos are just a tiny slice (less than 1% of all YouTube links AI Overviews actually cite). With the rest of the videos, the situation could be very different.”Quick GuideContact us about this storyShowThe best public interest journalism relies on first-hand accounts from people in the know.If you have something to share on this subject, you can contact us confidentially using the following methods.Secure Messaging in the Guardian appThe Guardian app has a tool to send tips about stories. Messages are end to end encrypted and concealed within the routine activity that every Guardian mobile app performs. This prevents an observer from knowing that you are communicating with us at all, let alone what is being said.If you don't already have the Guardian app, download it (iOS/Android) and go to the menu. Select ‘Secure Messaging’. SecureDrop, instant messengers, email, telephone and postIf you can safely use the Tor network without being observed or monitored, you can send messages and documents to the Guardian via our SecureDrop platform.Finally, our guide at theguardian.com/tips lists several ways to contact us securely, and discusses the pros and cons of each. Illustration: Guardian Design / Rich CousinsWas this helpful?Thank you for your feedback.Explore more on these topicsGoogleAI (artificial intelligence)YouTubeHealthAlphabetInternetnewsShareReuse this contentMost viewedMost viewedUKUS politicsWorldClimate crisisMiddle EastUkraineFootballNewslettersBusinessEnvironmentUK politicsScienceTechGlobal developmentObituariesNewsOpinionSportCultureLifestyleOriginal reporting and incisive analysis, direct from the Guardian every morningSign up for our emailAbout usHelpComplaints & correctionsContact usTip us offSecureDropPrivacy policyCookie policyTax strategyTerms & conditionsAll topicsAll writersNewslettersDigital newspaper archiveBlueskyFacebookInstagramLinkedInThreadsTikTokYouTubeAdvertise with usGuardian LabsSearch jobsWork with usAccessibility settings Back to top© 2026 Guardian News & Media Limited or its affiliated companies. All rights reserved. (dcr)

Google’s AI Overviews, a tool generating summaries at the top of search results, are heavily relying on YouTube for health information, a study has revealed, raising significant concerns about the reliability of this widely used AI assistant. The research, conducted by SE Ranking in December 2025, analyzed over 50,807 healthcare-related queries in Germany and found that YouTube was the most cited source – accounting for 20,621 citations out of a total of 465,823. This surpasses even prominent medical reference sites like Msdmanuals.com and reputable German consumer health portals.

The study’s findings directly challenge Google’s assertion that AI Overviews prioritize “high-quality content from reputable sources,” regardless of format. Instead, it highlights a structural vulnerability—a reliance on a platform known for its vast, largely unvetted content, where anyone can upload information, including individuals with no formal medical training. The researchers emphasize that this situation isn’t an isolated incident, but a systemic issue with implications that could extend beyond Germany.

The research underscores the critical distinction between popularity and reliability when assessing health information. While YouTube boasts a massive user base, its content lacks the rigorous editorial oversight and adherence to evidence-based standards characteristic of established medical institutions and organizations. The study documented this by revealing the sources that the AI Overviews utilized. The most cited domain was YouTube with 20,621 citations out of a total of 465,823. The next most cited source was NDR.de, with 14,158 citations (3.04%). The German public broadcaster produces health-related content alongside news, documentaries and entertainment. In third place was a medical reference site, Msdmanuals.com with 9,711 citations (2.08%). The fourth most cited source was Germany’s largest consumer health portal, Netdoktor.de, with 7,519 citations (1.61%). The fifth most cited source was a career platform for doctors, Praktischarzt.de, with 7,145 citations (1.53%).

Google stated that AI Overviews surface high-quality content. However, the researchers argue that this is largely driven by visibility and popularity, rather than by medical accuracy. They point out that of the 25 most cited YouTube videos, 96% were from medical channels, but these represented only a tiny fraction (less than 1%) of all YouTube links cited by AI Overviews. Despite this, the situation could be very different.

The research’s findings come on the heels of a Guardian investigation that revealed the potential for harm when Google’s AI Overviews provided inaccurate and misleading health information. A particularly alarming example involved false information about liver function tests, potentially leading individuals to underestimate the severity of their condition. Google subsequently removed the AI Overviews for some, but not all, medical searches, highlighting the precariousness of the situation.

Hannah van Kolfschooten, a researcher specializing in AI, health, and law not involved in the study, corroborated these concerns. She stated, "This study provides empirical evidence that the risks posed by AI Overviews for health are structural, not anecdotal. It becomes difficult for Google to argue that misleading or harmful health outputs are rare cases." Instead, she explained that the findings demonstrate that these risks are embedded in the design of AI Overviews, driven by a preference for visibility and popularity over medical reliability.

Google defended its approach, stating that AI Overviews were designed to surface high-quality content regardless of format and that the most cited domains were indeed reputable websites. However, this defense does little to mitigate the fundamental problem: the reliance on a platform with inherent vulnerabilities. The study clearly demonstrated that Google’s AI Overviews are susceptible to incorporating misinformation and unverified claims, potentially putting public health at risk. The research underscores the urgent need for Google to reassess its approach to AI-powered search summaries and to prioritize accuracy and reliability over algorithmic popularity.