Hundreds of private conversations with Anthropic's artificial intelligence chatbot Claude were publicly accessible through Google search results, exposing personal information, CVs and proprietary corporate research.
The links to the chat logs, some containing users' names, contact details and work history, appeared in search results when a site-specific search term was used. The chats were from users who had chosen to "share" a link to their conversation, which search engines then indexed and made available to the broader public.
“Hundreds of private Claude AI chats exposed via Google after users shared links.”
Users on Reddit initially discovered the publicly available chats, which covered more than 200 conversations across at least 25 pages of search results, some taking place only weeks ago.
In one conversation from April, a user prompted Claude to draft an unpublished blog post about cloud security involving details of a corporate project. Another user asked Claude how to "become become Nine-tailed fox?" and the chatbot initially responded with an AI-generated image claiming they had been given "fully functional fox powers!"
Some users conducted what appeared to be proprietary research for their work, such as in healthcare, including transcripts of private conversations. Others sought help with their CVs, sharing their names, contact information and work history.
Anthropic removed the search availability of the chat logs over the weekend, but many had already been saved and shared widely online.
A spokeswoman for Anthropic said that Claude users maintained control over if and when to share conversations. She said links were "not guessable or discoverable unless people choose to share them themselves".
"When someone shares a conversation, they are making that content publicly accessible, and like other public web content, it may be archived by third-party services," the spokeswoman added.
The share option within Claude tells a user that "anyone with the link" may view the contents, but does not explicitly state that the link may end up in Google search results.
The incident mirrors a similar issue with OpenAI's ChatGPT last year, when chat logs were made publicly accessible, prompting the company to change how easily such logs could be found.