An unknown number of Claude chats and Artifacts — the interactive mini apps and documents that users can create inside Anthropic’s AI assistant — were found to be publicly searchable on Google over the weekend. Reddit users reported that Google search operators such as site:claude.ai/share surfaced a long list of shared conversations. Some of the results reportedly included health records, private company documents, and the names and phone numbers of children.

How the exposure occurred

The exposure appears to have stemmed from Claude’s “share chat” feature, which lets users generate links that allow anyone with the assigned URL to view a conversation or project. In Claude’s interface, the feature warns users: “Anyone with the link can view.”

That wording indicates the feature is meant for sharing with friends, colleagues, or limited groups, rather than the broader internet. Google Docs has a similar link-sharing option, but those documents do not automatically become publicly accessible through Google Search.

The distinction is important because public-link sharing typically controls access through possession of the URL, not through user authentication. If that URL is later posted on a public website, indexed by a search engine, or archived by a third-party service, the content can become visible far beyond the original intended audience.

Claude’s interface includes one option to “Keep private” and another to “Create public link.”

Anthropic’s response

Anthropic appeared to place responsibility for the exposure on users. Asked what happened, the company told Bitcoin World that share links appear in search results only when they have been posted somewhere search engines can access, such as a forum or social media post. Anthropic added that a link sent privately to another person remains outside search.

Anthropic spokeswoman Amie Rotherham said in an explainer: “We give people control over sharing their Claude conversations publicly, and in keeping with our privacy principles, we do not share chat directories or sitemaps with search engines like Google. These shareable links are 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 issue was first flagged by a Reddit user on Saturday and was first reported by 404 Media on Monday morning. By Monday afternoon, a Bitcoin World test search on Google using the method described in the Reddit post did not return any results, suggesting the exposure had somehow been remediated.

What was exposed

Before the issue was fixed, Futurism reported finding “a detailed medical report of a real patient, clinical trial results that included patient names, documents sharing the names and phone numbers of primary school-aged children, company documents marked for internal use only, and employee reviews that included personal information about workers.”

The exposed Artifacts included code, work notes, and erotica generated by the chatbot. In at least one case, Fortune reported that a chat labeled “shared by Anthropic” showed Claude producing explicit content, which would violate Anthropic’s own policy against generating erotica.

Previous incidents

Last year, Forbes reported a similar incident in which hundreds of Claude chats were indexed by search engines. At the time, Google estimated that it had indexed just under 600 conversations before the pages disappeared from search results.

The scale of the current exposure has not been independently confirmed, although multiple users reported finding shared conversations through the same type of Google search query used to uncover last year’s cache. Also last year, 404 Media reported that a researcher was able to scrape around 100,000 ChatGPT conversations that had been set to public sharing.

Google’s position

Google spokesperson Ned Adriance told Bitcoin World: “Neither Google nor any other search engine controls what pages are made public on the web, and these pages were indexed across many search engines. We give site owners clear controls to decide whether pages can be crawled or indexed, and we always respect those directives.”

Search engines generally rely on website operators to use standard crawling and indexing controls, such as directives that tell crawlers whether a page should appear in search results. Google’s statement frames the Claude pages as public web content once the links were available to crawlers, while Anthropic’s statement says the company did not provide chat directories or sitemaps to search engines.

How to check shared Claude chats

Claude users can review which chats they have set to public link sharing by going to Settings → Privacy → Shared Chats. That section allows users to see which conversations have been shared and revoke access if necessary.

Why the incident matters

The incident underscores a recurring privacy challenge involving AI chatbots: the gap between what users expect and how visible shared content can become. Anthropic’s interface warns that shared links are public, but many users may not realize that posting a link on a public forum, social media platform, or even a semi-public site can make the content discoverable through search engines.

As AI tools become more widely used in work and personal contexts, the risk of accidental data exposure increases. Users should treat any shared AI conversation link as a public record that could be archived or indexed, even when the interface frames the link as a tool for limited sharing.

FAQs

Q1: How did Claude chats end up on Google?

Users shared their Claude conversations through the “share chat” feature, which creates a public URL. When those URLs were posted on forums, social media platforms, or other public websites, search engines such as Google indexed them and made them discoverable through search.

Q2: Is my data safe if I never shared a Claude chat?

If a user never used the “share chat” feature, their conversations are not publicly accessible through that mechanism. The exposure affected only chats that users explicitly shared through public links.

Q3: Has the issue been fixed?

As of Monday afternoon, searches using the method described in the Reddit post no longer returned results, suggesting the exposure had been remediated. Users should still review their shared chats under Settings → Privacy → Shared Chats to confirm that no unintended public links remain active.