An untold number of private Claude AI chats and interactive Artifacts containing sensitive medical records, corporate data, and children’s contact details were exposed on Google Search over the weekend after users discovered that shared links were being crawled and indexed by search engines.
How Claude’s Share Feature Exposed Sensitive Data
The security loophole stems from Claude’s “share chat” tool, which generates unique URLs allowing users to share their conversations or interactive projects with others. Although the user interface states that “anyone with the link can view”, the phrasing strongly implies the tool is intended for controlled sharing among colleagues or friends, rather than public indexing. Unlike similar document-sharing services like Google Docs, which keep shared links hidden from search engines, Claude’s shared links found their way onto public search results.
Anthropic Attributes Exposure to User Sharing Habits
Anthropic has deflected blame onto its users for the search engine visibility. When questioned about the leak, the AI startup told TechCrunch that shared links only become searchable if they are posted on public platforms, such as forums or social media feeds, where search engine crawlers can detect them. The company maintained that links shared privately remain unindexed.
Anthropic spokesperson Amie Rotherham clarified the company’s stance, stating: “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.”
Highly Sensitive Data Discovered in Search Results
The privacy issue was initially flagged by a Reddit user on Saturday, with 404 Media first reporting the leak on Monday morning. While a follow-up test search on Monday afternoon suggested the exposed links had been removed from Google’s index, the brief window of exposure revealed highly confidential information.
Before the links were removed, Futurism reported discovering a trove of highly sensitive data. This included detailed medical reports of real patients, clinical trial results featuring patient names, internal corporate documents, employee performance reviews, and the full names and phone numbers of primary school-aged children.
Bypassing Safety Filters: Claude Erotica Exposed
The leak also exposed interactive Artifacts containing code, working notes, and even instances of the AI generating sexually explicit content. According to a report by Fortune, at least one indexed chat labelled “shared by Anthropic” showed Claude producing erotica.
While Anthropic’s official safety guidelines strictly prohibit Claude from generating sexually explicit material, users have frequently found creative ways to bypass these guardrails across various AI platforms. It remains unclear how the user in question prompted Claude to generate the erotica, and Anthropic has not yet commented on this specific incident.
Google Responds as History Repeats for AI Chatbots
Google spokesperson Ned Adriance addressed the indexing issue, stating: “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.”
This is not the first time Anthropic has faced indexing issues. Last year, Forbes highlighted a similar incident where search engines indexed hundreds of Claude chats. During that leak, Google estimated it had indexed around 600 conversations before they were removed. ChatGPT faced a comparable issue last year when a researcher scraped approximately 100,000 ChatGPT conversations that had been set to public sharing.
How to Secure Your Claude Conversations
If you have used Claude’s sharing feature and want to review or revoke public access to your chats, navigate to Settings -> Privacy -> Shared Chats within the Claude interface to manage your active links.