Meta finds its Muse

Meta’s new Muse AI agent is just over two weeks old and already speed-running its way past the industry’s biggest names.

Less than two weeks after launching, the AI agent had already climbed to the top of the free app leaderboards on Apple’s App Store, ahead of ChatGPT, Gemini, Claude, and other top AI tools.

So, what is Muse? Meta’s new app features a highly polished interface—centered around an oddly-adorable digital assistant—where users can give it tasks to carry out in plain English, like booking travel, sending emails, or shopping across the web.

Meta is also testing a voice feature that allows Muse to make phone calls on a user’s behalf, offering a glimpse at a future where AI assistants increasingly act as real-world stand-ins for their humans.

Playing to their strengths

With Muse, analysts say Meta is hoping to capitalize on its built-in advantages like product design, massive distribution, and a deep understanding of consumer behavior, rather than attempting to roll out the smartest underlying AI model.

Meta is able to market Muse to billions of people through products which are already integrated into their daily lives—like Facebook and Instagram—while also tapping into an ecosystem of accounts, contacts, and personal context that could make AI agents more useful from Day 1.

But…There is a catch: the more Muse does on your behalf, the more trust it requires, including access to your email, calendar, financial accounts, contacts, and other key personal information.

This could prove a particularly high hurdle for Meta given the company’s lengthy history of privacy and safety controversies, though Zuckerberg has emphasized Muse was “built from the ground up for privacy and security.”

Looking ahead…AI-powered shopping could soon become a new battleground for retailers, as a wave of AI agents promise to simplify users’ lives. On Sunday, Amazon blocked Meta’s Muse from accessing its website, saying the “unauthorized AI agent” violated its terms of service.