Meta has introduced Muse, an AI agent capable of autonomously handling complex tasks such as booking travel, selling items, and sending emails by accessing various apps on behalf of users. This launch marks a pivotal step in CEO Mark Zuckerberg's vision to provide personal superintelligence to billions of Meta platform users.
- Muse can autonomously access multiple apps to perform complex tasks.
- The product will launch first in the US with free and subscription options.
- Meta has implemented robust safety protocols amid mixed internal test results.
What happened
Meta launched its AI agent Muse (internally called Hatch) in the United States, allowing users to delegate tasks such as sending emails, booking travel, and making payments by connecting Muse to their apps. The agent is available via a standalone Muse app and Meta's WhatsApp messaging platform. Meta plans to extend Muse's availability to smart glasses in the near future.
Muse operates in the cloud on virtual machines that emulate personal computers, enabling it to carry out requests continuously in the background. Meta offers a basic free version of Muse and two paid subscription tiers priced at $20 and $100 per month for heavier usage. Users can opt out of their data being used for AI model training, with encrypted versions expected later.
Why it matters
Muse represents the cornerstone of Meta's strategy to pivot beyond advertising revenue and capitalize on its vast investments in AI infrastructure projected to exceed $130 billion this year. By offering personal superintelligence through Muse, Meta aims to deepen user engagement while creating diversified revenue streams from AI-based subscriptions.
Given Muse’s deep integration with sensitive personal data spanning email, calendar, health, payments, and smart home devices, security and privacy risks are elevated. To mitigate these, Meta has embedded separate monitoring agents that oversee Muse’s actions and require authorization before certain operations, reflecting the company's heightened focus on safety following internal delays and testing challenges.
What to watch next
Despite its potential, internal feedback has been mixed, with some employees praising Muse's utility during activities like managing vacation logistics, while others identify critical security vulnerabilities and reliability issues. Observers will be looking for how effectively Meta addresses these concerns in upcoming software updates and the promised encrypted version.
Future developments to monitor include Muse’s integration into Meta’s hardware ecosystem such as smart glasses and the rollout of more sophisticated security features. Additionally, market adoption rates of paid subscriptions and any regulatory scrutiny over privacy implications will be key indicators of Muse's commercial and ethical success.