Claude Makes Five AI Labs Publishing Private Chats To Google

Anthropic became the fifth major AI lab to have shared chat pages surface in Google Search, exposing some Claude conversations that contained sensitive medical and personal information.

The indexed pages were removed from search results within about 48 hours, but the underlying share links could still be opened directly.

The incident highlights a broader privacy risk: AI chat share features can unintentionally turn private conversations into public web pages.

It also raises questions about how well safeguards like robots.txt and X-Robots-Tag work for preventing accidental indexing.