‘Dangerous and alarming’: Google removes some of its AI summaries after users’ health put at risk
Exclusive: Guardian investigation finds AI Overviews provided inaccurate and false information when queried over blood tests
TL;DR
- Google has removed AI Overviews for specific liver function test search queries due to inaccurate and potentially harmful information.
- The Guardian investigation highlighted how the AI summaries provided misleading context for liver test results, risking patient harm.
- Health charities and patient advocates remain concerned about the broader accuracy and potential dangers of AI-generated health information on Google Search.
- Google stated they continuously review and improve AI Overviews, citing internal clinician reviews that found many instances of information to be accurate and supported by reputable sources.
- Concerns persist that AI Overviews might still appear for other health-related queries with "completely wrong" and "really dangerous" information, including cancer and mental health topics.