What People Suggest: The Short Life of Google’s Most Controversial Health AI Feature

A health AI feature from Google that crowdsourced medical advice from internet users has been discontinued, leaving behind questions about how the company manages responsible deployment of AI in healthcare. “What People Suggest” organized user-generated health content from online discussion platforms using AI and presented it as part of search results for health queries. Three insiders confirmed its removal, and Google acknowledged the decision following a journalistic inquiry.
The product was introduced at Google’s annual “The Check Up” event in New York, with then-chief health officer Karen DeSalvo highlighting the value of peer health experiences for users managing long-term conditions. The AI-organized community content was displayed thematically alongside links to original sources. The feature was rolled out to mobile users in the US.
The company’s explanation for removing the feature — search interface simplification with no safety concerns — was criticized when the blog post cited as public disclosure contained no mention of the discontinued feature. The lack of a clear public communication has been characterized as emblematic of how Google manages inconvenient product decisions.
The context includes a significant investigation earlier this year that found AI Overviews on Google Search were providing inaccurate health information to approximately two billion users monthly. Google’s selective rollback of medical AI Overviews following that finding was seen as inadequate. The issue of AI health misinformation on the world’s most used search engine remains live and contested.
The next iteration of “The Check Up” is expected to showcase fresh AI health research and partnerships. For these announcements to land with credibility, Google will need to be willing to engage openly with past failures — including the handling of “What People Suggest.” That kind of honest engagement is, ultimately, what trustworthy health AI requires.