
What happened
Meta launched Muse, a free U.S. personal agent that searches products, navigates checkout and pays through Stripe's Link after the shopper approves the total.
Why it matters
AI-assistant traffic to retail sites jumped 693.4% in the 2025 holiday season, yet just 31% of Vogue Business respondents said they would outsource shopping to an agent.
What to watch
Muse's payoff may hinge on whether shoppers accept Meta's data safeguards, given its 2019 record $5 billion FTC privacy penalty. Watch the opt-out controls.
WHO IT HITSRetail and e-commerce teams weighing whether to open their checkout flows to AI agents, and privacy and ad-tech staff at Meta who must show that Muse's data handling differs from past practices, are the roles most directly affected.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
Meta's Muse joins a crowded field: Amazon's Alexa for Shopping in May, Disney's conversational shopping assistant tests in June, and Uber Eats' Cart Assistant that turns a prompt or photo into a cart. Meta's own path here was costly. After Llama 4 was widely panned, Meta paid $14.3 billion for a 49% stake in Scale AI, brought in cofounder Alexandr Wang as chief AI officer, and committed hundreds of billions of dollars to AI infrastructure. Muse runs on Muse Spark, which the company's own results place in the race without consistently beating OpenAI, Anthropic, and Google.
The consumer picture Muse walks into is mixed. AI-assistant traffic to retail sites rose 693.4% during the 2025 holiday season and online holiday spending hit a record $257.8 billion, up 6.8% from the previous year, per Reuters. But in the Vogue Business survey, 72% said they would not share card details and only 2% said AI consistently understood their personal style—suggesting shoppers want help narrowing options, not an agent deciding for them.
Meta's edge is memory: years of what users watch, follow, save and share. The company says users control which apps connect, can disconnect anytime, and can opt out of AI training, and that Muse conversations are not shared with its ad systems. Those safeguards must contend with a record $5 billion FTC penalty in 2019 over Facebook privacy allegations. Whether Muse becomes useful or intrusive likely hinges on whether shoppers accept that trade-off—and on Meta respecting the opt-out controls it has promised.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Much of the attention on AI infrastructure buildouts is now tied to sheer compute power, with dominance define…

Barron's reported September 10 that Kepler Computing emerged from stealth with a memory architecture using fer…

Dynatrace acquired Arize AI, adding AI observability, evaluation and agent monitoring to its application obser…
Reuters reported September 10 that inference-chip startup d-Matrix will use Nvidia's NVLink Fusion to connect…

A Daily Dose of Data Science test kept LoRA adapters separate from a shared 7B base model, cutting 100 fine-tu…

A report by Spencer Kitts, Thomas Larsen and Sydney Von Arx says an OpenAI agent swarm very likely ran an atta…
