All NewsEducationTVBrokers
Equities & FundsCrypto & Digital AssetsAI & TechnologyBusiness & CorporateUS Politics & PolicyGeopolitics & Global RiskMacro, Rates & FXCommodities & EnergyEuropean Politics & MarketsAsia-PacificReal Estate & Property
All NewsHome
← Back to AI & Technology

Meta launches AI agent that can access other apps to send emails, make payments

Created at 8 Sep · 7:07 PM2 sources↑ Market-relevant2 events
IN SHORT

Meta has introduced Muse, a new personal AI agent designed to perform everyday tasks by connecting to users' apps and services, including email, calendars, and payment platforms. The agent aims to move beyond simple question-answering to actively completing tasks for users.

Key Numbers

$18 billionmultistate settlement over social media harms
$20/monthPower subscription plan for Muse
$100/monthMaximum subscription plan for Muse
2011year Meta settled with FTC over privacy charges
$5 billionFTC penalty in 2019 over privacy violations
2019year Meta discovered user passwords in readable formats
$942 millionpayment ordered in New Mexico lawsuit over child harms

Who's Involved

Meta
company that launched the Muse AI agent
Muse
Meta's new personal AI agent
Muse Spark
Meta's AI model powering the Muse agent
Stripe
payment platform integrated with Muse for checkouts
Shopify
company whose Shop Pay integration is coming to Muse
1Password
company whose integration is coming to Muse
FTC
U.S. agency that has previously penalized Meta for privacy violations
Mark Zuckerberg
CEO of Meta Platforms Inc.
Meta launches AI agent that can access other apps to send emails, make payments

↳ Why This Matters

Meta's launch of Muse signifies a major push into agentic AI, potentially reshaping how consumers interact with technology for daily tasks. However, the company's past privacy issues present a significant hurdle to widespread adoption, creating a critical test of consumer trust in AI services that require deep personal data integration.

Key facts

  • Meta launched Muse, a personal AI agent that connects to users' apps and services.
  • Muse can perform tasks like sending emails, booking travel, filling out forms, and making purchases.
  • The agent is powered by Meta's Muse Spark AI model and uses built-in connectors or user-provided credentials for app access.
  • Muse will be available for free with paid subscription plans offering increased usage.
  • Meta states Muse operates in a secure virtual machine to protect user data and privacy.
  • Meta has a history of privacy violations, which may impact consumer trust in Muse.

Meta has launched Muse, its most significant consumer AI offering to date, aiming to move beyond simple chatbots to an AI agent capable of performing real-world tasks. The agent connects to users' existing apps and services, such as email, calendars, and payment platforms, to execute actions like sending emails, booking travel, and making purchases.

Muse is powered by Meta's Muse Spark AI model and features built-in connectors for various services, with plans to integrate more over time. Users can also set up connections for services with public APIs or access them via a browser if no API is available. The opt-in nature allows users to connect apps one at a time, enhancing transparency.

The agent will be accessible through its website, iOS and Android apps, WhatsApp chats, and Meta's AI glasses. While free to use, Meta will offer two paid subscription plans, Power ($20/month) and Maximum ($100/month), for users who require more extensive task delegation. A usage meter will inform users about their remaining free usage.

Meta emphasizes Muse's security, stating it operates within a dedicated, secure virtual machine called Muse Secure VM, isolated from Meta's advertising systems. A separate Sentinel agent runs on the same VM, ensuring Muse cannot access user passwords or payment methods. Despite these assurances, Meta's history of privacy violations, including significant FTC settlements and past data breaches, raises concerns about consumer trust.

The company is attempting to rebuild trust by providing detailed technical documentation and allowing users to customize Muse with a name and avatar, fostering a sense of personal connection. Whether these measures will be sufficient for consumers to entrust Meta with more personal information remains to be seen.

Frequently asked questions

Muse is a new personal AI agent from Meta designed to help users with everyday tasks by connecting to their apps and services like email, calendars, and payment platforms.

Muse can send emails, book travel, fill out forms, create plans, generate grocery lists from recipes, send invitations, and make purchases.

Meta states Muse operates in a secure virtual machine (Muse Secure VM) and does not share user conversations or data with Meta's ad systems. It also claims Muse cannot see user passwords or payment methods.

Muse is free to use, with two paid subscription plans available: Power at $20/month and Maximum at $100/month, offering increased usage for task delegation.

Meta has a history of privacy violations, including FTC settlements and past data breaches, which may lead consumers to question the security of their personal information with the new AI agent.

What Happens Next

01Meta plans to add more service integrations to Muse over time.
02Muse will soon be available through Meta's AI glasses.
03Security experts will investigate Meta's claims about Muse's data protection measures.
CME Headlines
  • Risk Management and Monitoring Notice: Multi-Factor Authentication Updates - September 12
    3 Sep · 5:00 AM

How It Developed

Meta launched its AI assistant, Muse, which can access apps to perform tasks like sending emails and booking travel, despite internal data security concerns.
Meta introduced Muse, a personal AI agent that connects to users' apps and services to perform tasks such as sending emails, booking travel, and making purchases.
Muse is powered by Meta's AI model, Muse Spark, and offers built-in connectors for various services, with plans to expand integration options.
The agent will be available via web, iOS and Android apps, WhatsApp chats, and Meta's AI glasses, with free usage and paid subscription tiers.
Meta claims Muse operates within a secure virtual machine, Muse Secure VM, with a separate Sentinel agent to protect user data and prevent access to passwords or payment methods.
Despite security claims, Meta's history of privacy violations and data mishandling raises questions about consumer trust in the new AI agent.

Sources

T1
Meta launches AI agent that can access other apps to send emails, make paymentsReuters
T1
Meta debuts its Muse AI agent. Will consumers trust it?TechCrunch

Related Stories

Apple's Ternus faces AI test with new foldable iPhone and Siri overhaul
8 Sep · 10:10 AM
AI assistant books travel, orders groceries, and handles customer service
8 Sep · 6:26 PM
Goldman Sachs opens AI engineering hub in Bellevue
8 Sep · 3:06 PM
US Accuses Chinese AI Firms of Industrial-Scale Technology Theft
8 Sep · 5:52 PM
Eric Wu's NavigateAI targets construction labor shortage with AI copilots
8 Sep · 2:41 AM