Meta Pauses Muse Link-Sharing After Researchers Find Thousands of AI Chats Exposed in Public DNS Records
Security firm Claroc found over 6,800 Muse conversation links through public DNS records, exposing chats about revenue, loans and workplace calendars. Meta confirmed the issue and paused sharing in October 2026 while reviewing link discovery.
Meta has paused a sharing feature in its Muse AI assistant after security researchers discovered that links to user conversations were publicly accessible through DNS records. The company confirmed the pause in October 2026 following a report by security firm Claroc.
Claroc researchers found more than 6,800 Muse share links by searching public DNS records — a system that functions as a directory for web traffic. The firm explained that Meta constructed each share URL using the conversation’s title, a short descriptive phrase, and a random ID. Because the hostname embedded that same information, researchers could identify and reconstruct full links without ever opening them first.
Among the conversations Claroc was able to view were a marketing company’s revenue review, a small business owner’s loan application, a workplace’s full-year staff calendar, and notes from a job candidate’s call with an AI lab.
Muse spokesperson Tracy Clayton confirmed the issue and the company’s response in a statement to The Verge: “People have the ability to share specific Muse artifacts, which creates a page that can be accessed by those with the link. However, we’ve paused this functionality while we review this report and consider changes to how these links can be discovered.”
The Muse sharing feature was designed to let users send a link to their AI conversation with another person. The problem, as Claroc identified, was that the URL structure made those links discoverable by anyone searching public DNS records — not just the intended recipients.
This is not the first time AI chat-sharing features have exposed conversations more broadly than users may have intended. The source notes that conversations from OpenAI’s ChatGPT, Anthropic’s Claude, and xAI’s Grok have previously appeared in Google Search results, suggesting the issue of unintended AI chat exposure extends across the industry.