RismadarVoice Reporters
September 9, 2026
Meta has launched a new artificial intelligence assistant designed to perform tasks on behalf of users, including sending emails, making payments, booking travel and interacting with other applications.
The AI agent, called Muse, was introduced on Tuesday as part of Meta CEO Mark Zuckerberg’s plan to develop what the company describes as “personal superintelligence” for users of its platforms.
Meta said Muse will initially be available in the United States through a dedicated application and its WhatsApp messaging service. The company also plans to integrate the assistant into its smart glasses in the future.
The basic version of Muse will be available for free, while subscription plans costing $20 and $100 per month will be offered for users requiring heavier usage.
The company said users will be able to connect Muse to applications including email, calendars, payment services, health platforms, shopping accounts and smart home systems. Users can choose which apps the AI can access and withdraw permission at any time.
Meta said it delayed the launch earlier in the year to improve security measures, adding that the system was designed with safeguards to reduce risks associated with handling sensitive personal information.
However, internal testing of the AI agent reportedly raised concerns about reliability and privacy. Some Meta employees reported instances where the system failed to complete tasks properly or bypassed certain restrictions during testing.
One internal report cited by Reuters said Muse exposed personal iCloud photos after being prompted to identify objects in images from a child’s birthday party. Other testers reported issues including frequent logouts, failed monitoring tasks and unexpected errors.
Meta Vice President of AI Products Vishal Shah said the company understands that mistakes may occur but believes the technology has met the required safety standards for public release.
The launch comes as major technology companies continue expanding AI agents capable of independently carrying out digital tasks. However, concerns remain among experts about privacy, security and the risks of giving artificial intelligence access to personal accounts and sensitive information.
Meta has invested heavily in artificial intelligence infrastructure, with the company expecting its AI-related spending to exceed $130 billion this year as it seeks to expand beyond its traditional advertising business.
