Google’s AI Overviews are increasingly the first place people turn for health information, but new analysis suggests the sources behind many of those summaries fall short of traditional medical standards.
The issue is less about opposing AI and more about whether accountability is keeping pace as automated tools shape health decisions.
Recent reporting by The Guardian has renewed scrutiny of Google’s AI-generated health advice, citing warnings from medical charities and experts. Examples included flawed dietary guidance for pancreatic cancer and misleading explanations of liver blood test results, raising concerns about the real-world consequences of inaccurate summaries.
Google has pushed back on the criticism. The company said the examples were “taken out of context”, argued that “most AI Overviews are accurate”, and maintained that its summaries “link to reputable sources”. Still, questions remain about how those sources are selected.
A new study by SEO platform SE Ranking shifts the focus from individual answers to the wider system behind them. The analysis reviewed 50,807 health-related searches in Germany, examining where Google’s AI Overviews draw their information.
“Nearly two-thirds of Google AI Overview citations come from sources without strong medical or evidence-based safeguards.” That points to a structural issue rather than isolated errors.
YouTube emerged as the single most cited source, accounting for 4.43 percent of all health-related AI citations. Only 34.45 percent came from more reliable medical sources such as hospitals, clinics and health associations. Academic journals and government health institutions together made up around 1 percent.
The study also found a disconnect between AI answers and traditional search results. YouTube ranked first in AI citations but only 11th in organic search, while just 36 percent of AI-cited pages appeared in Google’s top 10 results.
“Google should be held to the same standard it has long imposed on others” when it comes to health information.