Meta introduces Muse: The new personal AI agent ready to run your life

From booking holidays to negotiating bills, Meta’s latest innovation promises to be the ultimate digital personal assistant
Meta launches secure Muse AI agent to automate your life
Mark Zuckerberg
Updated on
2 min read

Meta has officially launched Muse, a personal AI agent designed to autonomously handle everyday tasks and long-term goals. Developed as part of CEO Mark Zuckerberg’s vision for "personal superintelligence", the tool operates via a dedicated app and WhatsApp. Mark anticipates that this new agent will allow people to focus on what truly matters to them. From booking travel and negotiating bills to transforming a saved Instagram recipe into a tailored grocery list, Muse promises to take time-consuming projects off your plate.

Meta launches secure Muse AI agent to automate your life

Unlike standard chatbots, Muse is built to execute real-world tasks independently in the background. It features a built-in browser and integrates with Stripe Link to generate single-use payment cards, allowing users to shop safely across the internet while keeping their real financial details hidden. The basic programme is free for up to 100 million tokens per week, with premium subscriptions priced at $20 and $100 a month for heavier usage. The company also plans to integrate the agent into its line of smart glasses soon.

Meta introduces Muse: The new personal AI agent ready to run your life
Meta launches secure Muse AI agent to automate your lifeUnsplash

Given its access to sensitive personal data, security is a major priority. The agent operates within an isolated cloud environment called Muse Secure VM. A separate monitoring system known as Sentinel ensures that nothing reaches the internet without strict approval. Vishal Shah, vice president of AI products at Meta, delayed the initial release to improve safety. Vishal noted that while perfection is impossible, "every single part of the architecture has been designed to make this as safe, as secure, as private as we can possibly make it".

Despite these ambitious safeguards, internal testing has revealed some teething issues. Employees have reported mixed results, with chief technology officer Andrew Bosworth highlighting technical glitches. Andrew noted he was repeatedly logged out during use, while other testers flagged problems with the agent stalling when monitoring fast-selling tickets and bypassing guardrails to access private photos. To address ongoing vulnerabilities, Meta is offering rewards of up to $300,000 for researchers who find security flaws.

For more updates, join/follow our WhatsApp, Telegram and YouTube channels.

Meta launches secure Muse AI agent to automate your life
Someone filming you without consent with Meta smart glasses? Internet says, play a Disney song

FOLLOW US

ON GOOGLE DISCOVER

X
IndulgExpress
www.indulgexpress.com