Meta's Muse Agent Hits 5 Million Downloads, Lifts Stock 17%
Meta's Muse agent passed 5 million downloads in two weeks and pushed Meta stock up 17%. Wall Street is now repricing retailers built on consumer inertia as autonomous AI handles purchasing decisions.
Updated
Why it matters
- Muse passed 5 million downloads since its September 8, 2025 release, with more than 3 million weekly active users
- Meta shares rose more than 17% from launch through Thursday's close
- Since Muse's debut, Planet Fitness dropped 12%, Booking Holdings fell 11%, Schwab lost 9%, and SiriusXM dropped 8.6%
- TJX gained nearly 8%, Costco rose 4%, and Amazon slipped 1% over the same window
- FactSet projects Amazon 2025 advertising revenue at $83 billion, while Amazon said Alexa for Shopping users spend 40% more per order on average
Meta Platforms' personal artificial intelligence agent (a software system that executes tasks on a user's behalf) Muse has passed 5 million downloads since its September 8 release, and Meta shares have climbed more than 17% on the launch — a stock reaction that pushed agentic AI, defined as autonomous tools that execute multi-step tasks without continuous user input, into the center of retail-sector positioning on Wall Street.
Muse displaced OpenAI's ChatGPT at the top of Apple's U.S. App Store charts within days of release. More than 3 million users now run the software weekly. The agent searches retail catalogs, compares prices across sites, and completes purchases on behalf of consumers.
What does the launch change for retailers?
Investor reaction to Muse has fractured the consumer-discretionary universe into winners, losers, and what Goldman Sachs calls the "consumer inertia trade" — companies that profit when customers delay cancelations, skip price comparisons, or stay subscribed by default. Since Muse's debut through the prior Thursday's close:
- Planet Fitness fell almost 12%
- Booking Holdings dropped 11%
- Charles Schwab lost 9%
- SiriusXM dropped 8.6%
Goldman Sachs framed the shift as a move from "search and shop" to "delegate and approve," a transition that hands transactional decisions to software rather than to human shoppers. Current AI shopping tools still focus on product discovery and recommendations. Goldman expects those tools to capture more of the purchasing workflow in coming quarters.
Three retailers — TJX, Costco, and Amazon — have reacted differently. TJX gained nearly 8%. Costco rose about 4%. Amazon slipped 1%. The S&P 500 added roughly 1.7% over the same window. Each company carries a distinct defense against agent-mediated commerce.
What is Amazon's exposure?
Amazon, the largest U.S. e-commerce platform, looks like an obvious target. An external agent could route orders past Amazon's search bar and toward competing marketplaces. The company has responded by blocking Muse from completing purchases directly on its site, even as Walmart, Shopify, and Best Buy have permitted agent traffic.
"Amazon and [other] companies that have that strong value proposition are less exposed," Rosenblatt Securities analyst Scott Devitt said. Goldman reached a comparable conclusion, arguing that "platforms offering compelling prices, broad inventory selection, fast fulfillment, trusted consumer relationships, and rich first-party data assets are likely to be best positioned."
Amazon has not stood still. The company's Alexa for Shopping assistant served more than 350 million customers over the trailing twelve months, according to CEO Andy Jassy on the most recent earnings call. Active users nearly doubled in the second quarter. Interactions increased more than fivefold year over year. Amazon stated that customers using Alexa for Shopping spend 40% more per order on average than non-users.
How does advertising complicate the picture?
The deeper pressure point sits in Amazon's advertising business, which FactSet projects at $83 billion in 2025 revenue. Sponsored listings depend on shoppers searching, browsing, and clicking within Amazon's interface. As agents route decisions elsewhere, that revenue stream narrows.
Telsey Advisory Group analyst Joe Feldman argued that an outside agent could insert distance between Amazon and its shoppers even when Amazon physically fulfills the order. Amazon would keep the sale. Amazon would lose the first-party data and the brand impression that travel with a direct visit.
Devitt expects Amazon to find alternative revenue paths — agent-tier advertising, higher seller fees, or Prime repricing. Two metrics to track: advertising revenue growth, and any expansion or compression of sales and marketing expense.
Could AI validate Costco's value proposition?
Costco runs on membership fees, low prices, and impulse-driven store trips. An agent programmed to execute a fixed shopping list would strip out the in-aisle discovery that drives unplanned purchases. The same tool could also confirm Costco's prices against competitors in real time.
CFO Gary Millerchip told investors on the most recent earnings call that sales originating from large language model (LLM) searches — Gemini, Anthropic's Claude, and OpenAI among them — are growing at triple-digit rates, though from a small base. He framed the trend as a validator of membership value rather than a threat.
"AI presents an opportunity in a sort of neutral environment for us to be able to ensure that our value and our quality shows up transparently for our members," Millerchip said.
The dynamic cuts both ways. Agents could intensify price sensitivity across the category and turn the executive membership fee into a constant comparison. Costco's fiscal 2026 fourth-quarter results showed renewal rates improving and executive memberships reaching an all-time high in penetration. Executive members renew at higher rates than basic members.
The wrinkle is demographics. Costco is gaining younger shoppers. Younger members sign up online at higher rates. Online sign-ups carry higher churn (the rate at which customers cancel or fail to renew) than in-store sign-ups. Retention metrics for those newer cohorts become a primary indicator as agent tools spread.
Why might TJX be harder to disrupt?
TJX's T.J. Maxx and Marshalls run a "treasure hunt" model built on a constantly rotating, partially unsearchable assortment. Off-price retail, or the discount sourcing of canceled production orders and excess manufacturer inventory, relies on product going to stores rather than to websites. Discounts are not advertised widely online.
"You really can't find an agent to find treasure," Wells Fargo analyst Ike Boruchow said. "You're like, 'I need some clothes, and I'm going to go into TJX and see if I find anything that's fun.'"
Guggenheim analyst Simeon Siegel sees consumer behavior bifurcating into needs and wants. Agents will absorb routine household replenishment and subscription cancelations first. "There is something special. There is something personal. There is subjectivity around buying discretionary items, around picking out a fall sweater, around choosing your sneakers," Siegel said.
"I think off-price wins because they don't have e-commerce, not in spite of it," Siegel added. "How does an agent help you find something that is not searchable in the first place, how does an agent help you shop somewhere whose intention is to sell invisibly through a store?"
TJX confirmed the point on its own earnings call. Executives conceded that merchandising missteps in the Marmaxx segment, the unit that houses T.J. Maxx and Marshalls, produced softness in the most recent quarter. CEO Ernie Herrman told investors a fix has been deployed. The episode underscored that even insulated models depend on consistent in-store execution. "I think it'll be a tool to shop, not the exclusive way to shop," Siegel said.
What investors should watch next
The Muse launch has reset the stakes for retailers built on consumer inertia. Three signals will mark the pace of change: agent adoption rates beyond the early-adopter cohort, the persistence of LLM-driven sales lift at Costco, and whether Amazon's Alexa for Shopping keeps converting agent users into higher-value shoppers.
Goldman frames the next phase as "delegate and approve." The market is currently pricing agentic commerce as a slow roll. A single quarter of faster adoption could reweight those stock charts overnight, and the retailers built around invisible inventory, validated membership value, or walled-garden scale will set the pace for everyone else.
Source: CNBC Tech
More from Marcus Bennett
Show full bio
Senior reporter covering companies, deals and strategy in artificial intelligence and machine learning at AI In Context.
219 articles