Agentic commerce, where AI agents find and buy products on a person's behalf, is getting closer. Meta says it is developing a protocol that will give businesses more insight into how its Muse agents interact with their websites, with the aim of getting more sites to let Muse in.
As more AI agents crawl the web for users, many websites are blocking them. The concerns include agents probing for vulnerabilities, overloading systems or finding ways around restrictions and limits.
A human can't refresh a page 500 times a minute to grab a new product drop, but an agent can, and thousands of agents arriving at once could crash a site.
Meta is building the protocol with Sierra and partners across retail, payments and customer experience. According to Meta, it will handle authentication, empower consumers and give companies visibility into what personal agents do through their websites, APIs or company agents, and it will be open for anyone to implement.
The idea is to let websites run their own bots and controls that talk to Muse agents directly, so owners feel safer allowing agents to act for customers. The Muse app has passed 5 million downloads and 3 million weekly users.
Security is only part of it. If more transactions run through bots instead of people, brands could lose direct contact with customers, lose upselling opportunities and see the website display-ad model weaken. That's another reason some companies block agents. For privacy concerns around Muse, see What Personal Data Does Meta's Muse AI Collect? Wired Report Points to Hourly ‘Pages’ on Friends and Family.
Meta is already adding commerce features such as in-chat checkout to its ad tools; see Meta’s New AI Ad Tools at Advertising Week: A Business Assistant That Edits Budgets and In-Chat Checkout.
If you run a direct-to-consumer site, decide whether to block AI agents and under what conditions to allow them. Blocking everything means missing customers who shop through agents; allowing everything puts pricing, inventory data and servers at risk. Start by isolating agent traffic in your logs.
Agents don't see banners; they read product details, stock, prices and shipping terms. Accurate, well-structured product data is the baseline for agentic commerce.
An open protocol that handles authentication and consumer controls when personal AI agents like Muse interact with business websites, APIs or company agents, and shows businesses what those agents do.
Concerns about vulnerability probing, server overload and bypassed limits, plus fears of losing customer relationships, upsells and display-ad revenue.
Sierra and partners across retail, payments and customer experience; Meta says anyone will be able to implement it.
To apply what you just read to your own site, start with a free audit of where things are now.
A strategist replies within 24 hours on business days.
Naver’s shopping AI agent lifted transaction value 82% since June. Here’s how to prepare for agentic commerce: structured product attributes, agent-readable product pages, clear reasons to buy and trust signals.
Shipping delays, order cutoffs and fulfillment updates all need different timing and audiences. Here’s a simple framework for e-commerce customer messaging — manual, recurring and automated sends — plus SMS fallback.
Korean D2C brand Wiselly now sells every product at cost for a ₩2,990 monthly membership. How cutting distribution, marketing and packaging costs built its zero-margin commerce model.