Anthropic's Claude chats indexed, private data exposed

Hundreds of private conversations between users and Anthropic’s Claude AI chatbot were discovered to be publicly accessible through Google and other search engines over the past week, exposing personal and professional information that users believed was private. The searches showed Claude chats for which a user had decided to “share” a link had been saved by search engines like Google, leaving them accessible to the broader public. The search availability of the chat logs was removed over the weekend, but many were saved and shared widely online. This marks the second time Claude has experienced this issue, following a similar incident in 2025.

Locked padlock with digital breach showing exposed emails and data files, illustrating cybersecurity vulnerability and pri...

What Happened and How Chats Became Public

The issue involved conversations shared through Anthropic Claude’s public link feature rather than private chats being accessed through a security breach. Journalists discovered numerous Claude conversation pages indexed by search engines after using targeted search queries. The indexed pages reportedly contained a variety of user-generated conversations, including coding discussions, business planning, resumes, and other prompts that had been shared using public URLs.

The searchable conversations originated from Anthropic Claude’s sharing feature, which allows users to generate public links for individual chats. Once a conversation was shared through a public URL, search engine crawlers could discover and index the page if no technical measures prevented it. WIRED reported that thousands of shared Claude conversations were indexed before the issue received public attention.

Claude AI chat shared publicly, indexed by Google and Bing search engines, exposing private conversation data in search re...

After hitting the “share” button in the top right corner, Claude notes that sharing will create a “public” link, which “anyone with the link can view.” The warning, though, doesn’t alert users that shared content could wind up being indexed by a search engine. This meant users who thought they were sharing conversations with specific colleagues or associates inadvertently made them available to the entire internet.

Types of Content Exposed

The exposed conversations included sensitive personal and professional information that users likely did not intend to share publicly. The materials varied widely in nature, from business strategy documents to personal data. Many of the materials appear to have been made public when Claude users opted to distribute a given chat or project to colleagues or associates using the chatbot’s “share” feature. Other materials were shared in the form of Claude Artifacts, a feature that allows users to save certain projects as standalone apps or tools.

Exposed sensitive data including API keys, passwords, resume, business plans, and private messages on smartphone screen.

Notably, that platform does warn users that publishing an Artifact “will make it accessible to anyone on the internet and potentially visible in search engine results.” However, the standard chat sharing feature lacked such a clear warning about search engine indexing.

Claude warning messages about web search inaccuracy and public link sharing risks for private conversations

Anthropic’s Response and Technical Fix

A spokeswoman for Anthropic said that Claude users maintained control over if and when to share conversations they had with the chatbot. She said links to conversations were “not guessable or discoverable unless people choose to share them themselves”. The company maintained that the system was working as designed, emphasizing that only deliberately shared chats were affected.

Following the reports, Anthropic confirmed that Claude conversations remain private by default and that only chats intentionally shared through public links were eligible to appear online. The company also stated that it had implemented changes designed to prevent shared conversation pages from being indexed by search engines moving forward.

Diagram showing how to block search bots from indexing Claude conversation URLs using noindex meta tags and robots.txt dis...

As the search indexing of the chat logs is no longer occurring, it is likely Anthropic used available tools to quickly block the chat log links from search results. However, at the time of writing, many of the previously exposed chat links remained live and reachable to those already with the link.

Not an Isolated Incident in AI Industry

This privacy issue is not unique to Claude. When OpenAI last year experienced an almost identical issue with ChatGPT chat logs being made publicly accessible, the company ultimately changed the ease with which such logs were accessible. Grok, the AI chatbot within X, the social platform owned by Elon Musk, also last year saw hundreds of thousands of chat logs made publicly available though online search.

This isn’t the first time this has happened at Anthropic: Forbes reported last year that hundreds of Claude chats had been indexed by search engines, prompting Anthropic to scrub the chats from the web. The recurring nature of these incidents across multiple AI platforms raises questions about whether companies are doing enough to protect user privacy by default.

Search Engine Role and Responsibility

A spokesman for Google made clear to the BBC that the company does not control “what pages are made public on the web,” saying instead that action comes from websites. “We give site owners clear controls to decide whether pages can be crawled or indexed, and we always respect those directives.”

The indexed pages appeared across multiple search engines, not just Google. Google spokesperson Ned Adriance told TechCrunch that “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.”

Key Facts

Hundreds to thousands of shared Claude conversations were indexed by Google, Bing, and other search engines before the issue was addressed
The exposed content included coding discussions, business plans, resumes, and other potentially sensitive information
Search availability was removed over the weekend, but many conversations remain accessible to those with direct links
This is the second time Claude has experienced this issue, following a similar incident in 2025
OpenAI’s ChatGPT and Elon Musk’s Grok experienced nearly identical problems in 2025
Anthropic has implemented technical changes to prevent future shared conversations from being indexed