In a nutshell: Public chat links on Claude were indexed by Google and displayed in search results for months before Anthropic resolved the issue.
Anthropic’s Claude included public links to user chats in Google’s search index for an extended period, making shared conversations potentially discoverable to third parties. The company has now fixed the indexing problem.
Claude users have the ability to create conversation links and share them with others. However, these share links were indexed by Google’s search engine over an extended period, meaning they appeared in search results and were potentially discoverable by unauthorized parties.
The search engine indexing of chat links presents a security concern, as it made personally identifiable data such as conversation content, context, and potentially sensitive information publicly viewable. For CISOs, this represents a loss of control over data flow: users who believed they were sharing links only with selected recipients had no control over secondary distribution through the search engine.
Anthropic has remedied the issue following disclosure and adjusted the robots.txt file or similar control mechanisms to prevent further indexing of such links. Existing entries should also be removed from the Google index. This underscores the importance of privacy-conscious configuration in AI applications, particularly when users leverage features for collaborative work with potentially sensitive content.
Source: www.golem.de · Published 28 July 2026
Lumi AI News — AI-assisted curation in accordance with Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.7.3.