Public Claude chats appear in Google search, exposing sensitive user conversations

Google briefly indexed an unknown number of publicly shared Claude conversations over the weekend, allowing anyone to find them through simple search queries. The discovery raised fresh concerns about how AI chat-sharing tools handle sensitive information. Users reported finding conversations that included medical records, internal business documents, and personal details involving children. The issue centered on Claude’s chat-sharing feature. It lets users create a public link so others can view a conversation or an Artifact, Anthropic’s interactive documents and mini applications built inside Claude. Although the feature warns that anyone with the link can access the content, many users appeared to assume those links would stay limited to people they shared them with directly. Search results disappeared Reddit users first noticed the issue after using Google search operators targeting Claude’s public sharing pages. The searches surfaced numerous shared conversations that had become searchable through Google’s index. By Monday afternoon, those search results no longer appeared during tests. That suggests the exposed pages were removed from Google’s searchable index, although neither Google nor Anthropic explained exactly how the issue was resolved. Anthropic told TechCrunch that users control whether conversations become publicly accessible. “We give people control over sharing their Claude conversations publicly,” spokesperson Amie Rotherham said. Anthropic added that it does not provide chat directories or sitemaps to search engines. Rotherham said shared links cannot be guessed or discovered unless users publish them. She added that conversations become public web pages once someone chooses to share them. Like other public content, they may then be archived by third-party services. Anthropic also said conversations shared privately through a direct link do not appear in search results unless someone later posts that link somewhere search engines can crawl. Sensitive data exposed Before the pages disappeared from search, users reportedly uncovered a wide range of sensitive information. That included medical records, company documents marked for internal use and files containing children’s names and phone numbers. Some exposed Artifacts also contained software code and employee work notes. The findings renewed questions about whether users fully understand the visibility of AI-generated work shared through public links. Google said search engines only index content that website owners make publicly accessible. Google spokesperson Ned Adriance told TechCrunch that neither Google nor other search engines decide which pages become public. He added that Google provides website owners with controls to prevent crawling or indexing and respects those instructions. The incident echoes a similar situation reported last year, when hundreds of Claude conversations briefly appeared in search engine results before disappearing. Researchers also identified publicly shared AI conversations on other chatbot platforms during the same period.Users who have created public Claude links can review them through the Privacy section of Claude’s Settings menu under Shared Chats. The episode serves as another reminder that information placed behind a public sharing link can sometimes reach a much wider audience than expected.Recommended ArticlesGet the latest in engineering, tech, space & science - delivered daily to your inbox.Aamir is a seasoned tech journalist with experience at Exhibit Magazine, Republic World, and PR Newswire. With a deep love for all things tech and science, he has spent years decoding the latest innovations and exploring how they shape industries, lifestyles, and the future of humanity.