Agentic commerce flips online retail on its head. For two decades, the goal was to get a human to your product page and convince them to click buy. Now the buyer might be an AI agent that compares options, picks one, and checks out on a person’s behalf. If your store is invisible to that agent, you lose the sale before a human ever sees your brand.
TL;DR
- Agentic commerce is online shopping where an AI agent handles discovery, comparison, and checkout for the customer.
- Major platforms have built the plumbing fast, including OpenAI’s Agentic Commerce Protocol with Stripe and Google’s Universal Commerce Protocol.
- ChatGPT crossed roughly 900 million weekly users in early 2026, though OpenAI launched Instant Checkout in September 2025 and then pulled it back in March 2026, a sign even big platforms are still finding their footing.
- Trust is the brake. Around a quarter of consumers say they will never let an AI buy for them, so this is a shift to prepare for, not panic over.
What happened
The infrastructure for AI-driven buying came together quickly. OpenAI built an Agentic Commerce Protocol with Stripe, with partners including Instacart, DoorDash, Shopify, and Etsy. Google launched a Universal Commerce Protocol in January 2026 so agents can read merchant catalogues and complete purchases through one standard.
The reach is already large, and the path is not smooth. ChatGPT crossed roughly 900 million weekly users in early 2026. OpenAI launched Instant Checkout in September 2025, then discontinued it in March 2026 after a rocky start, even as it kept building agentic commerce plumbing with partners like PayPal and Stripe. As of March 2026, agentic storefronts are available to millions of Shopify merchants, and Walmart wired its own assistant into ChatGPT with plans to expand to Gemini.
Demand is forming, with caution attached. Surveys reported by payment firms suggest about a third of consumers expect at least 10% of their purchases to be AI-driven within a year. At the same time, 27% say they trust no organization to run an AI shopping agent, and 24% say they will never delegate a purchase. The road is being paved faster than shoppers are walking down it.
Why it matters in Canada
Canadian retailers and service businesses sit downstream of platforms built elsewhere, which means these protocols arrive whether or not anyone here votes for them. A Shopify merchant in Halifax can be included in agent-driven shopping by a decision made in a product meeting in another country. That is opportunity and exposure at the same time.
The mechanics reward structured, machine-readable information. An agent comparing products does not admire your brand photography. It reads your price, availability, specifications, shipping options, and reviews. Stores with clean, accurate, well-structured product data get surfaced. Stores that hide key details inside images or vague copy get skipped, because the agent cannot parse what it cannot read.
For Canadian sellers, there is a practical wrinkle too. Shipping times, duties, and currency are exactly the details an agent weighs, and Canadian operators who present that information clearly have an edge over those who bury it at checkout.
Business impact
The shift changes who you are selling to. Some of your future customers are software acting for a person, and software does not respond to a clever banner. It responds to data it can compare. That rewards retailers who treat their product feed as a core asset rather than an afterthought.
It also compresses the funnel. When an agent returns one recommendation instead of a page of options, being second-best is the same as being invisible. The margin for a weak listing shrinks, because there is no browsing step where a human might stumble onto you anyway.
The trust gap is the saving grace. With a quarter of consumers refusing to delegate purchases, this will not replace human shopping overnight. That gives Canadian businesses time to get their data house in order before agent-driven buying becomes routine for higher-value goods.
What leaders should do next
- Clean up your product data. Make sure price, availability, specifications, shipping, and returns are accurate and machine-readable, not locked inside images.
- Check your platform’s agent features. If you sell on Shopify or a similar platform, find out what agentic commerce options exist and how to turn them on or off deliberately.
- Make Canadian details explicit. State shipping times, duties, and currency clearly, since these are exactly what an agent compares.
- Strengthen reviews. Agents lean on ratings and third-party signals, so a healthy review profile improves your odds of being the one recommendation.
- Decide your policy. Choose whether you want agent-driven sales now, and set the terms before a platform default decides for you.
The skeptic’s view
A grounded retailer will say this is mostly platform marketing. Real shoppers still like to see, touch, and choose, and a quarter of them flatly reject the idea of an AI buying for them. Pouring effort into pleasing shopping agents that few customers use yet could mean neglecting the storefront real humans actually visit.
That is fair, and it argues against treating agentic commerce as today’s emergency. The counter is that the data work it requires, clean listings and honest shipping details, improves your normal store too. You are not betting the business on agents. You are doing housekeeping that pays off either way and leaves you ready if adoption accelerates.
What to watch
- Whether Canadian shoppers adopt agent-driven buying at the pace platforms expect, or whether the trust gap holds.
- How payment networks and regulators handle disputes, refunds, and liability when an agent makes a faulty purchase.
- Whether agent recommendations start carrying paid placement, which would change which merchants get surfaced.
- Platform defaults, since the moment your storefront is opted into agent sales matters as much as the feature itself.
Closing analysis
The retailers who win agentic commerce will be the ones whose product data is clean, honest, and easy for a machine to read. You do not need to chase every protocol. You need a store an agent can understand and a deliberate choice about whether you want its business. Do the data work now, while human shoppers still give you room.
FAQ
What is agentic commerce? It is online shopping where an AI agent handles discovery, comparison, and checkout on a customer’s behalf, rather than a person browsing and buying directly.
Is agentic commerce actually being used? The infrastructure is being built fast, with stumbles. OpenAI launched ChatGPT Instant Checkout in September 2025 and discontinued it in March 2026, while agentic storefronts reached millions of Shopify merchants by the same time. Many consumers remain hesitant to delegate purchases.
How do retailers stay visible to AI shopping agents? Keep product data clean and machine-readable, state shipping and currency details clearly, and maintain a strong review profile, since agents compare structured information rather than visuals.
Sources
- J.P. Morgan, “Agentic Commerce: The Future of AI-Powered Shopping.” https://www.jpmorgan.com/payments/newsroom/agentic-commerce-ai-future-shopping
- Checkout.com, “Consumer demand for AI shopping is forming fast but trust for agentic commerce is still catching up.” https://www.checkout.com/newsroom/consumer-demand-for-ai-shopping-is-forming-fast-but-trust-for-agentic-commerce-is-still-catching-up
- Shopify, “Agentic Commerce: Benefits and How To Get Started (2026).” https://www.shopify.com/blog/agentic-commerce
Related reading
- How to get your business recommended by AI search
- What agentic AI means for Canadian operators
- Most Canadian companies use AI and almost none make money from it
Disclosure
The author has no relevant financial, advisory, or board relationships with any party named in this column.
