Meta launches Muse personal AI agent as TechRadar questions its grip on real work

According to TechRadar's account of the launch, Muse is pitched as a combined work and home assistant that requires no technical expertise. Meta says it can handle tasks ranging from recipes and shopping lists to work duties such as calendar management, and that it ships with dedicated apps and a website.

Meta chief executive Mark Zuckerberg has said the company needs to build more accessible agents, with tools such as OpenClaw too advanced for most of Meta's users across Facebook, Instagram and WhatsApp.

On work tasks, Meta says Muse can draft and send emails, though the agent asks for approval before sending. The company describes wider ambitions, including "turning long-term goals into action plans." Meta says the agent can handle complex tasks independently and work behind the scenes even when the app is turned off. "Once a person shares a goal with Muse, it helps them develop a personalized plan and coordinate their time and resources, then advances the work on its own," Meta said in its launch post.

Meta presents Muse as an always-on assistant and co-worker, spanning holiday planning and fitness goals through to starting a business.

Security and privacy features figure prominently in the launch. Meta says personal agents like Muse "need a new kind of secure computer, so Meta built one for everyone," pointing to a Muse Secure VM that it says carries "first-of-its-kind privacy, safety, and security protections." Users decide how much access the agent receives, according to the company: "each person stays in control of their Muse and decides how much access it gets."

Muse can also set up connections to third-party services where a public API is available. TechRadar named Stripe, Google Workspace and 1Password among the examples cited by Meta, describing the arrangement as both a potential efficiency gain and a security risk.

Meta acknowledged the agent will not be flawless. Its launch post said: "Muse can and will still make mistakes, but we expect they'll be much less frequent and cause much less damage due to the safety systems we've built in."

TechRadar's opinion piece, whose headline described Muse as "a work accessory designed by people who don't do real work," questioned whether the feature list reflects how most people live. It said tools such as monitoring a smart home or planning a holiday may suit a Silicon Valley employee with an hour-long commute, but for people outside that environment "it's all a bit much." The piece also asked whether an agent that advances projects on its own removes the sense of achievement, and how far user control over data extends once daily life and work projects are fed into the system. It noted that the emphasis on privacy comes amid "the current furore around its AI 'Pervert glasses'." TechRadar concluded that the point of such technology should be to make life better, and that Meta "clearly need to talk to some actual people first."