The company has reactivated the ability for users to share content via its Muse AI chatbot after patching a security vulnerability. Meta’s current fix resolves a design flaw where URLs for shared artifacts were discoverable via passive DNS, which had previously prompted the company to disable the sharing feature to prevent unintended data exposure. All items shared before the patch have been assigned new URLs, meaning previous links are no longer functional.
The vulnerability was highlighted by security firm Claroc, which found more than 6,800 exposed links including bank fraud alerts, loan applications, and confidential company documents. Meta clarified that the flaw specifically affected shared artifacts rather than general chat sharing. To encourage responsible disclosure, the company maintains a bug bounty program for Muse that provides awards of up to $300,000 for valid findings.
