Meta says more than 10 million people in the US and Canada have activated its Muse AI agent, but executives at many businesses say agent-driven commerce has yet to make much of a mark. That gap between user interest and measurable sales is the useful signal: adoption counts are not the same as a shopping revolution.
Meta AI Watch analysis
What happened
Meta CEO Mark Zuckerberg said on the company’s earnings call that more than 10 million people had activated Muse, and that over a quarter of them used it for shopping. Meta says it has partnered with 10 retailers, including Walmart, Best Buy and Gap, to make shopping through Muse easier.
In a WIRED report published on 31 October, executives at around 50 companies described current agent-driven activity as small. DoorDash put orders from Muse and other agents in the “hundreds”; T-Mobile said it had seen nothing significant, while Hilton called bookings insignificant. Fitness chain Life Time Group said it had seen no cancellations prompted by agents.
There are early signs of businesses adapting anyway. Chipotle’s digital chief said Muse’s launch brought a single-digit percentage increase in agentic orders, and that the company is tidying its website code to help agents complete purchases. Amazon and Insurify, both of which had blocked Muse, have also changed their positions, according to WIRED.
Why it matters
Muse may be getting people to try AI-assisted shopping, but the companies that would feel the effects are not yet describing a major shift in sales. That distinction matters when one headline number measures activation and another question is whether agents are actually completing meaningful purchases.
Businesses are already weighing how to accommodate agents, while also watching security, mistaken orders and the possibility that automated comparison shopping squeezes profits. For now, the commercial story looks more like preparation and small experiments than a retail takeover. The robots have reached the checkout queue; the till, apparently, has not started ringing much louder.
Our read
The most revealing figure here is not the number of people who activated Muse, but the contrast between that audience and the modest activity businesses describe. Meta’s figures show interest, not proof of a new shopping habit at scale. The sensible test is whether completed transactions grow, and whether retailers can make agents useful without handing them a blank cheque and the account password.
What to watch
- Whether Meta publishes further data on completed purchases, rather than activations or shopping interest.
- Whether retailers report a sustained rise in orders from Muse and other agents.
- How companies handle agent access, mistaken orders and fees for transactions.
Discussion spark: Should retailers optimise their sites for AI shopping agents now, or wait until those agents are sending meaningful numbers of completed orders?
Sources and evidence
- Source update (31 October 2026, 10:30 UTC)
- Source update (31 October 2026, 10:30 UTC)
not affiliated with, endorsed by, or operated by Meta