Meta Launches Muse AI Agent Amid Trust and Safety Debate

Meta launched its Muse personal AI agent after Chief Executive Mark Zuckerberg pushed the company to move ahead despite internal safety concerns, as the company seeks a stronger position in the artificial intelligence market against rivals developing similar tools.

The New York Times reported that Zuckerberg discussed Muse in August with Alexandr Wang, Meta’s chief artificial intelligence officer, and Nat Friedman, head of A.I. product. The company publicly introduced Muse on Sept. 8 after concerns emerged during testing, including a reported case where the system changed a user’s password without permission.

Meta has positioned Muse as a central part of its AI strategy, with the app reaching the top of Apple’s free app rankings, according to CNBC. The company is now focused on whether users will continue using the assistant after initial interest fades.

"The real question is trust," said Meredith Whalen, chief research officer at IDC.

Whalen said consumers may find AI agents useful for completing tasks, but giving such systems access to personal information remains a major challenge, particularly for Meta because of previous disputes over platform safety and privacy.

"Muse has an opportunity to be sticky with consumers because it's a task-completer, rather than just a chatbot," said Meredith Whalen, chief research officer at IDC.

CNBC reported that Meta is offering Muse access for free while studying user behavior and considering future pricing models. The company also offers subscription plans costing $20 or $100 depending on token usage.

Other technology executives said broader adoption of AI agents is still developing. ThoughtSpot CEO Ketan Karkhanis described enterprise use of the technology as being at an early stage.

"I would say that widely, enterprise-scale AI agent adoption is still page one, chapter one," said Ketan Karkhanis, CEO of ThoughtSpot.

Meta also faces questions about how Muse handles information about people connected to users. A University of St. Gallen research group said it found features that could create profiles of people who are not Muse users.

"The people described need not be Muse users and have no say in being profiled," researchers wrote.

Meta spokesperson Daniel Roberts said the system needs context about users and their contacts to complete tasks. He said Muse gathers information from public sources and details users choose to share.

"Muse gathers that based on public information and from what you've chosen to share, which is how it remembers the person who sent you an invoice is in fact the plumber who you previously hired to complete work in your bathroom or which flowers your spouse said they liked best," said Daniel Roberts, Meta spokesperson.

Analysts have also examined possible business opportunities from Muse, including subscription revenue and potential effects on Meta’s advertising business. A Yahoo Finance opinion cited a Deutsche Bank analyst estimate that Muse could contribute up to 8% of Meta revenue by 2030.