Google has quietly removed several search results from its AI Overview feature after reports surfaced that it was providing misleading information and incorrect medical advice. Specifically, the tech giant has taken action to eliminate inaccurate guidance related to liver disease and liver function tests, ensuring that users no longer see AI-generated summaries for these sensitive queries.
Google quietly removes inaccurate health summaries
According to a report by Android Authority and an investigation by The Guardian, Google stripped certain medical information from its AI Overview following a wave of complaints. While health-related information remains accessible through standard Google Search, AI-generated summaries for specific liver-related phrases have been disabled.
It is worth noting that Google integrated AI Overview into its search engine last year to provide faster, more concise answers. However, the move to automate medical advice has proven controversial due to the high stakes of health-related data.
The danger of misleading medical results
When a user enters a query into Google, AI Overview typically generates a summarized response with links to relevant websites. However, The Guardian recently described some of these medical results as "extremely sensitive and dangerous."
A primary example involved liver function tests (LFTs). The AI-generated summaries provided inaccurate reference ranges that failed to account for critical biological factors such as:
- Age and Sex
- Ethnicity and Nationality
- Specific Lab Methodologies
Experts warn against "Dr Google" AI
Medical experts have labeled these AI-generated results as dangerous, warning that they could provide false reassurance. Inaccurate summaries might lead individuals with serious liver conditions to believe they are healthy, causing them to forgo necessary medical consultation or life-saving treatment.
In response to these findings, a Google spokesperson stated that the company is continuously striving to improve its policies and refine the accuracy of its AI models. Despite these removals, many other medical advisories continue to appear in AI Overviews.
