Users have discovered that some Claude conversations shared through public links were appearing in Google and Microsoft Bing search results, exposing discussions that included sensitive political, legal, and personal topics.

The incident highlights an important privacy issue: creating a shareable AI conversation link does not necessarily mean its contents will remain hidden from search engines, particularly if a platform does not implement all available indexing protections.

Public Links Exposed Sensitive Conversations

According to a report by WIRED, Claude allows users to generate public links that can be shared with others, creating a viewable copy of a conversation.

However, some of these shared pages became searchable through Google and Bing. At the time the report was prepared, approximately 612 Claude conversation pages were indexed in Bing.

Although Anthropic, the company behind Claude, used a robots.txt file to discourage search engine crawlers from accessing shared pages, that measure alone was not sufficient to prevent page titles and URLs from appearing in search results.

Blocking Search Indexing Requires More Than robots.txt

The incident underscores that AI platforms need stronger safeguards than relying solely on robots.txt. Web experts recommend using noindex meta tags or equivalent HTTP response headers on individual pages to explicitly instruct search engines not to index shared content.

Google noted that website owners are responsible for specifying which pages should or should not be indexed. Following the discovery, some of the exposed Claude conversations were removed from search results.

A Reminder About Sharing AI Conversations

The episode also highlights a broader issue surrounding users' expectations of privacy when sharing AI conversations.

Many people assume that a share link is intended only for selected recipients—such as colleagues, friends, or collaborators—and may not realize that, without appropriate technical protections, the content could become discoverable through public search engines.

The incident serves as a reminder that users should avoid sharing AI conversations containing confidential or sensitive information unless they fully understand how the platform handles public links and search indexing.