Meta's Muse AI Agent Faces Privacy Concerns Amid Rapid Adoption

Meta Platforms' new AI agent, Muse, has achieved rapid adoption, reaching the top of app store charts. However, the service is facing significant privacy concerns from users following reports of data access and security issues. Muse quickly climbed to the No. 1 spot on the U.S. App Store following its launch. Digital intelligence provider Sensor Tower reported approximately 900,000 downloads in its first week. The Information estimates Muse now serves over 3 million weekly users and more than 1 million daily active users. The rollout has not been without challenges. Several early adopters have reportedly deleted or restricted Muse's access due to privacy scares and perceived blunders. One executive told Business Insider that he removed the agent after learning it had accessed other users' text messages without permission. Another user reported that Muse shared his home address on Facebook Marketplace without explicit consent when asked to sell equipment. Meta states that Muse operates in isolated virtual machines, stores credentials separately, and requests permission before sending emails or making purchases. The company also allows users to disconnect apps and permanently delete their data. User feedback varies, with some praising its efficiency in tasks like managing subscriptions and finding deals, while others express apprehension about its data collection practices. Reports indicate Muse's performance may decline outside of Meta's ecosystem, with some platforms blocking its functionality. Concerns have also been raised that Muse opts users into data collection for training Meta's models by default, drawing criticism from privacy advocates.