Google removes certain AI summaries related to health issues
What's the story
Google has removed some of its AI-generated health summaries after a Guardian investigation found that they contained false and misleading information, putting users at risk. The firm's AI Overviews, which use generative AI to provide quick snapshots of important information on a topic or question, were found to be inaccurate in some cases. These inaccuracies could lead to serious health risks for users.
Health risks
AI overviews lead to misinterpretation of liver tests
One particularly alarming case involved false information about normal ranges for liver blood tests. The AI Overviews provided a lot of numbers with little context, failing to account for factors like nationality, sex, ethnicity or age of patients. This could lead seriously ill patients to wrongly believe they had a normal test result and skip follow-up healthcare meetings.
Action taken
Google removes overviews for specific search terms
In light of the investigation, Google has removed AI Overviews for the search terms "what is the normal range for liver blood tests" and "what is the normal range for liver function tests." A company spokesperson said they don't comment on individual removals within Search, but work to make broad improvements when context is missed. They also take action under their policies where appropriate.
Ongoing worries
Concerns over potential misleading AI overviews
Despite the removal of some AI Overviews, concerns remain that potentially misleading information could still be provided if questions are asked differently. Vanessa Hebditch, Director of Communications and Policy at the British Liver Trust, said while they welcomed Google's action in these cases, their concern is that other AI-produced health information might be inaccurate and confusing.
Persistent issues
Google continues to face criticism
Google's AI Overviews still appear for other examples highlighted by The Guardian, including summaries about cancer and mental health. These have been described as "completely wrong" and "really dangerous." When asked why they had not been removed, Google said they linked to well-known and reputable sources, informing users when it was important to seek expert advice.