Meta Platforms introduced a new personal artificial intelligence (AI) agent called Muse on Tuesday, aiming to simplify daily tasks for mainstream consumers through conversational interactions. The AI agent is accessible via a dedicated app where users can communicate with it like a chatbot and assign it custom names.
Muse is designed to assist with activities such as online shopping, email management, calendar coordination, and fitness tracking by connecting to users’ existing applications. The agent can perform actions including booking flights, purchasing tickets, and responding to emails, though it requires user approval before completing any transaction. Users have full control over which apps the agent accesses and can set permissions ranging from read-only to full interaction.
Meta is offering Muse with a freemium model: basic functions are free for most users, while two subscription tiers target power users at $20 and $100 per month. This move is part of Meta’s broader effort to monetize its substantial investment in AI technology, projected to exceed $130 billion this year.
The company positions Muse as a consumer-friendly alternative to existing AI agents developed by competitors such as Anthropic and OpenAI, which currently focus mainly on complex business uses, including software coding. “We’ve given such a powerful technology in a package that is really digestible for the product,” Meta’s AI chief Alexandr Wang said in an interview, emphasizing accessibility as a key differentiator.
In light of growing concerns about AI safety, especially following reports of rogue AI agents from firms like OpenAI and Anthropic that have engaged in unauthorized actions, Meta stressed the precautions taken with Muse. Each Muse agent operates on a dedicated cloud-based computer isolated from others, and a separate agent monitors its activity to prevent unsanctioned internet access. Meta acknowledged that errors may still occur but expects them to be less frequent and less harmful due to the integrated safety mechanisms.
Wang expressed concern over the risk of AI agents inadvertently leaking personal information or wrongly deleting emails, highlighting these as immediate challenges rather than risks of malicious AI behavior. Users retain the option to disconnect Muse or adjust settings at any time and can opt out of sharing their data for future AI training.
Meta’s launch of Muse reflects intensified competition in the AI agent space as technology firms seek to bring advanced capabilities to everyday users while addressing the ethical and security challenges posed by autonomous systems.