PSA: Your Claude shared chats and Artifacts may have ended up on Google

Anthropic’s Claude, an AI system operated by the company Anthropic, has come under scrutiny after reports that shared chat links and Artifacts generated through the service may have been indexed by Google’s search engine, according to TechCrunch. The outlet reports that the exposure appears to have originated from Claude’s “share chat” feature, which lets users generate links granting anyone with the URL access to view a conversation or project.

The mechanism resembles issues seen previously with other shareable-link features across the tech industry, where content intended for limited distribution becomes discoverable through search indexing once a link is public and unguarded by access controls. TechCrunch’s report does not specify how many shared chats or Artifacts were affected, nor does it detail how long the exposure persisted before being noticed.

The report does not indicate what action, if any, Anthropic has taken in response, and TechCrunch’s account does not include comment from the company confirming the mechanism or scope of the disclosure.

Shared links of this kind are commonly used by Claude users to distribute conversation transcripts or generated projects to collaborators or the public. The incident underscores a recurring pattern in consumer software, where convenience features built for sharing carry unexamined implications for search visibility and inadvertent disclosure.