Story
September 14, 2026
Meta Wants Muse to Shop for You, but Consumers Aren’t Ready to Hand Over the Cart
Meta presents Muse as a controllable assistant that can turn its vast understanding of users into practical help, while consumers appear far more comfortable with AI narrowing choices than with a machine—or Meta—taking control of their money and data.
On Tuesday, Meta entered the accelerating AI-shopping contest with Muse, a free personal agent initially rolling out to U.S. adults. The company says it can handle tasks from messaging friends and planning evenings out to finding products and moving through checkout—then show the total for the user’s approval.1
That is a sharper proposition than a chatbot offering comparisons. Muse can draw on information users have shared across Meta’s services: a saved Instagram recipe reel, for example, can become a grocery list. Meta says people decide which apps the assistant connects to, can revoke access, and may opt out of having conversations used to train AI models. It also says Muse data will not feed its advertising systems.1
But the convenience pitch lands against an old and awkward backdrop: Meta’s record $5 billion FTC penalty in 2019 over allegations that Facebook violated users’ privacy. The more Muse knows, the more tailored its recommendations could become; for many shoppers, that exchange is precisely the problem.1
There is clearly demand for some AI help. Retail traffic from AI assistants and chatbots surged 693.4% during the 2025 holiday season, the article notes. Yet a Vogue Business survey found only 31% would outsource shopping to an agent that knew their tastes and purchase history. More pointedly, 72% said they would not share card details, while just 24% trusted AI-generated recommendations and summaries.1
Muse is also a high-stakes consumer test of Mark Zuckerberg’s AI reset after the poorly received Llama 4, including Meta’s $14.3 billion investment in Scale AI. Amazon, Disney and Uber are pursuing versions of the same idea, and Mastercard forecasts that one in 10 shoppers will use a personal AI agent by 2030.1
For now, the gap is simple: shoppers may welcome a tool that finds a cheaper TV or completes a tedious grocery list. They are less eager to let it decide what to buy. As the article puts it, the challenge is knowing “when consumers want an agent to take over—and when they merely want someone to hold the bags.”1