Tech News
← Home  ·  All topics

Claroc

3 GoKawiil briefs on this topic

Meta's Muse AI shared chat links that exposed users' private data

Security researchers at Claroc found that Meta's Muse AI agent generated shareable web links for user chat logs that could be discovered and opened by outsiders. Using a lookup dataset of visited URLs, investigators retrieved over 6,800 Muse share links in a single search, revealing private material such as corporate revenue documents, loan applications, bank transaction details, income data, and a workplace vacation calendar. Meta told Ars Technica it has disabled the sharing feature.

Meta pauses Muse AI chat-sharing feature after exposed conversation links found online

Meta has temporarily disabled a sharing feature in its Muse AI assistant after security firm Claroc reported finding more than 6,800 shareable conversation links through public DNS records. The links, built from conversation titles and random IDs, could reveal page contents through hostnames, exposing private material such as a company's revenue review, a loan application, and internal staff calendars. Meta spokesperson Tracy Clayton confirmed the pause, saying the company is reviewing the report and considering changes to how such links can be discovered.

Meta's Muse AI chatbot leaked shared conversations via guessable URLs

Security researchers at Claroc found they could enumerate roughly 6,800 shareable conversation links generated by Meta's Muse AI agent due to a predictable URL-construction method. Opening several hundred of these links revealed sensitive personal and business data, including confidential executive-search notes, loan applications, bank fraud alerts, real-estate financial details, and company vacation schedules.