Companies

Shopify opens checkout to browser-based AI agents

Shopify says browser-based AI agents can now complete purchases via new WebMCP checkout tools, while Amazon and Adidas block agent-driven buying on their platforms.

Shopify opens checkout to browser-based AI agents
Shopify opens checkout to browser-based AI agentsAI-generated
By James Calloway5 min read

Updated

Why it matters

  • Shopify announced Monday that browser-based AI agents can now complete purchases on merchants' sites via new WebMCP checkout tools: get_checkout, update_checkout, and complete_checkout.
  • Checkout support, including Shop Pay, lets agents read, update, and submit transactions with buyer authorization without screenshots or webpage scraping; the rollout covers all eligible merchants.
  • The tools run on Shopify's Universal Commerce Protocol (UCP); agent partners Muse and Instinct already have direct Shopify partnerships, with the Instinct deal announced the same day.

Shopify announced on Monday that browser-based AI agents can now complete purchases on its merchants' sites, extending agent capabilities beyond searching for products and adding items to carts. The move positions Shopify as the most aggressive major commerce platform in the emerging agentic shopping market, at a moment when other retailers are pulling in the opposite direction.

Amazon — and Adidas, apparently — are blocking AI agents from making purchases on users' behalf on their platforms, according to the report. Shopify has moved the other way, betting that structured machine access to checkout will drive sales rather than expose merchants to automated abuse.

From carts to completed orders

Shopify previously supported WebMCP for its storefronts and carts, allowing browser-based AI agents to comb through a Shopify retailer's inventory, search for products, and add them to carts. The checkout step remained the last mile: an agent could build a cart but could not close the transaction. That gap is now closed.

The update adds WebMCP support for checkout, including Shop Pay, the company said. That means agents can read the checkout screen, update it, and submit the transaction with the buyer's authorization — without relying on screenshots or scraping webpages.

The distinction matters. Screen-scraping and screenshot-based agents operate on interfaces built for human eyes, inferring meaning from rendered HTML. Structured APIs hand agents machine-readable data directly, which reduces errors, preserves required disclosures, and creates an auditable path for authorization.

Three new tools

The feature introduces three new tools:

  • get_checkout — allows an agent to inspect the current state of a checkout
  • update_checkout — lets it change things like the customer's address or delivery option
  • complete_checkout — places the order after the buyer authorizes it

The authorization requirement is built into the flow. An agent can read and modify a checkout on its own, but the final submission happens only after the buyer approves the transaction.

The feature is rolling out to all eligible Shopify merchants, said Gil Greenberg, a staff product manager who works on agentic commerce at Shopify, in a post on X.

Two protocols, one foundation

The checkout rollout fits into a broader architecture Shopify has been assembling for machine-driven commerce. The company already offers a hosted Model Context Protocol (MCP) server, which allows agents to work server-to-server. The proposed standard WebMCP, meanwhile, is designed for agents that work inside the buyer's browser.

Both leverage Shopify's Universal Commerce Protocol (UCP), which provides a common way to search for and discover products, build carts, and check out. In practice, that gives agent developers two integration paths — one where the agent talks to Shopify's systems directly, and one where it operates alongside the shopper in the browser — while relying on a single underlying protocol for commerce semantics.

Greenberg's guidance to developers was explicit about which path to use when. "If your agent is operating in the buyer's browser, use WebMCP tools provided on storefront and checkout to efficiently complete order placement, instead of navigating HTML built for humans," he wrote on X. "These WebMCP tools provide structured and efficient APIs, purposely designed — via UCP — to ensure accurate commerce facts, required disclosures, and handoff requirements."

The phrase "required disclosures" signals one of the sharper technical points in agentic commerce. Regulatory and commercial rules often mandate that buyers see specific information — shipping terms, taxes, refund policies — before completing a purchase. Agents parsing raw HTML can drop or mangle those disclosures. A protocol-level guarantee keeps them intact.

Partners are already live

Agent developers do not have to wait for the ecosystem to form around the new tools. Top AI agents like Muse and Instinct already have direct partnerships with Shopify for agentic commerce. The Instinct partnership was announced the same day as the checkout rollout.

That timing suggests Shopify is coordinating infrastructure and demand side by side: shipping the rails on Monday and announcing a named agent partner using them the same day. For merchants, it means agentic purchases are not a theoretical future capability but something that can begin flowing through checkout as the rollout reaches their stores.

Why the split with Amazon matters

The divide now running through e-commerce is not about whether AI agents will shop — it is about whether platforms will let them. Amazon's approach, blocking agent-driven purchases, defends a business built on owning the customer relationship, the interface, and the recommendation layer. Blocking agents keeps that funnel inside Amazon's own surfaces.

Shopify's calculus differs. It powers millions of independent merchants rather than operating a single storefront, and every additional sales channel — social, mobile, now agentic — extends the value of its platform. Opening checkout to agents makes merchants' inventory reachable from wherever the buyer's agent happens to live.

The risk is real on both sides. Platforms that block agents may lose sales to merchants who accept them, as buyers delegate more of the shopping process. Platforms that open up must guarantee that authorization is genuine, disclosures survive the transaction, and merchants retain visibility into what agents do on their sites. Shopify's answer — structured tools with buyer authorization as an explicit step in complete_checkout, and UCP-enforced disclosure requirements — is an attempt to encode those guarantees in protocol rather than policy.

What to watch

The rollout to all eligible merchants will be the first test of whether browser-based agents can handle real transactions at scale without the fragility of scraping. Watch how quickly Muse, Instinct, and other partners move from cart-building to full checkout flows, and whether merchants report meaningful order volume through the new tools.

The bigger question is competitive. If agentic purchasing takes off and Shopify is the place where it works, the company will have converted an AI-era threat — agents bypassing storefronts entirely — into a distribution advantage. Amazon's blockade, by contrast, is a bet that buyers will keep coming to human-facing interfaces. Monday's announcement means that bet is now being tested.

Original: bloomberg.com

Share this article:

More from James Calloway

James Calloway

Show full bio

News editor covering industry trends and analytics at AI In Context.

121 articles

Related articles

  1. AMD to acquire Fei-Fei Li's World Labs for $8.2 billion in stock
  2. Meta launches enterprise business to monetize its $600 billion AI bet
  3. OpenAI's Math Advisory Group Off to Another Rocky Start
  4. Meta Launches Enterprise AI Platform, Poaches MongoDB CEO to Run It

« Previous articleNext article »