Google has withdrawn some of its AI-generated health summaries after an investigation revealed that the tool was providing misleading and potentially dangerous medical information. The summaries, known as AI Overviews, appear at the top of search results and are designed to give quick explanations. However, in certain cases they presented inaccurate data that experts warned could put users at risk.
The investigation found that searches related to liver blood tests produced lists of “normal” values without sufficient context. Falling to consider factors such as age, sex, ethnicity or nationality. Health specialists cautioned that this could lead people with serious liver disease to believe their results were normal and skip essential follow-up care. Following these findings, Google removed AI Overviews for specific liver test queries, though similar summaries can still appear for slightly rephrased searches.
Health charities and patient information advocates welcomed the removals but stressed that deeper concerns remain. They argued that the issue goes beyond individual search terms and reflects broader risks associated with AI-generated health information, which can oversimplify complex medical issues and offer false reassurance. Critics also warned that AI Overviews often fail to clearly signal uncertainty or the need for professional medical advice.
Google defended the overall reliability of its system. Arguing that it reviews and improves AI Overviews and relies on reputable sources, with internal clinicians assessing disputed cases. However, experts and watchdogs maintain that errors in health-related summaries carry heightened consequences, especially given Google’s dominance in global search. The episode has intensified scrutiny of how AI tools handle sensitive health information and whether current safeguards are sufficient to protect public trust and safety.
Reference
Gregory, A. (2026, January 12). ‘Dangerous and alarming’: Google removes some of its AI summaries after users’ health put at risk. The Guardian. https://www.theguardian.com/technology/2026/jan/11/google-ai-overviews-health-guardian-investigation?CMP=Share_AndroidApp_Other