Google’s AI-powered health summaries are under intense scrutiny after a Guardian investigation uncovered several cases of misleading — and possibly even harmful — medical advice. The tool, called “AI Overview,” sits right at the top of Google search results and offers quick answers to complicated questions. While it’s supposed to make trustworthy health information easier to find, experts found the tool sometimes gives out incorrect or risky advice.
In one instance, the AI Overview gave wrong dietary tips to people with pancreatic cancer — advice doctors say could leave patients severely undernourished and unable to handle critical treatments. Other examples include the tool misreading liver blood test results and spreading false details about women’s cancer screening.
Medical professionals have voiced strong concerns, warning these errors could make people ignore important symptoms, delay seeking help, or even put themselves at risk by following unsafe advice. Charities and health experts also pointed out that these AI-generated summaries often lack proper context, and that the tool sometimes delivers inconsistent answers to the same question, making it difficult for users to trust the information.
Sophie Randall, who leads the UK Patient Information Forum, summed it up by saying Google’s AI Overview could “pose health risks by placing inaccurate health information at the top of online searches.” Anna Jewell from Pancreatic Cancer UK agreed, cautioning that the faulty nutrition advice could stop patients from getting enough calories, threatening their ability to manage treatment or surgery.
Google has responded, saying most health summaries produced by AI Overview are “accurate and helpful.” The company claims it is working hard to improve the quality of health-related information and insists that these summaries always link to trustworthy sources. They also advised users to reach out to healthcare professionals for any crucial medical advice.
The controversy raises serious questions about how reliable AI is for sharing health facts online. Experts say extra care and human checks are needed to prevent technology from causing real-world harm.
