Google pulls some AI health summaries after misinformation concerns

Google has removed certain AI-generated health summaries from its search results following reports that inaccurate information on medical tests could mislead users, renewing concerns over the reliability of artificial intelligence in delivering sensitive health-related content.

Google has taken down some of its AI-generated health summaries after concerns were raised that incorrect information could put users at risk. The move follows reports highlighting inaccuracies in Google’s AI Overviews, a feature that uses generative artificial intelligence to present brief answers at the top of search results.

The issue came to light after investigations found that certain AI Overviews provided misleading information about liver function tests. According to experts, incorrect ranges shown for these tests could cause individuals with serious liver conditions to mistakenly believe their results were normal, potentially delaying medical care. Health professionals described the errors as worrying, given the reliance many users place on search results for initial medical guidance.

In response, Google removed AI Overviews for specific queries related to liver blood tests and liver function test ranges. The company, however, declined to comment on individual changes made to its search results.

Health experts urge caution

While welcoming the removal of the inaccurate summaries, health organisations have warned that broader risks remain. The British Liver Trust said misleading AI-generated health content could still appear if questions are phrased differently, creating ongoing confusion for users seeking medical information online.

Google has acknowledged that AI Overviews may sometimes lack important context. A company spokesperson said Google continues to refine the feature and takes action when content falls short of its internal standards. The company added that its AI summaries are designed to appear only for queries where it has high confidence in the quality of information.

Balancing AI and medical responsibility

Despite the removals, AI Overviews continue to appear for other health-related searches, including topics such as cancer and mental health. Google said these summaries are supported by links to well-known and reputable sources and often include prompts encouraging users to seek professional medical advice.

According to Google, its internal team of clinicians regularly reviews feedback and evaluates the accuracy of AI-generated responses across different categories. The company said these reviews help improve reliability over time.

The episode has reignited debate around the use of AI in health information, with experts stressing that while technology can improve access to knowledge, it should not replace professional medical judgment.