Shopify will now pay AI agents to show your products. Neutral shopping answers just got harder to assume.

There is a sentence in Shopify's developer documentation that we think will matter more than most announcements this fall. It describes how an AI shopping agent gets products from Shopify's Global Catalog:

The catalog's server-managed configuration determines whether Shopify blends promoted placements into the ranked results.

Put plainly, an agent that searches Shopify's catalog can now receive a list where some products are promoted, and the developer of that agent earns a commission when those products sell. Shopify's changelog says the program, Promoted Placements, entered beta on October 27 and now includes merchants running Shop Campaigns.

This is not a scandal. Affiliate commissions are how a large part of the web pays for product reviews. But it changes a basic assumption many people hold about AI shopping, and brands should think it through before the holiday rush.

What Shopify built

The mechanics, from Shopify's documentation, are straightforward.

Table: Shopify Promoted Placements, as documented

Detail	What Shopify says
Who gets paid	Developers of agents and apps that search the Global Catalog and opt in
Base rate	0.3% of attributed purchases of eligible products
Higher rates	Merchants running Shop Campaigns can fund extra commission, added to the base rate
Where the higher rate applies	US, in USD, only when the buyer completes the purchase in Shop or on shop.com
Attribution	Last non-direct click, 7-day window
Scope	Every eligible item in the attributed order, not only the item clicked
Disclosure	Developers must disclose the paid relationship and label promoted placements
Status	Invite-led; rates and fields may change before general availability

Shopify's example shows a promoted product carrying a total commission of 1.5%. Products from merchants outside Shopify are not eligible and never carry a placement marker.

Why this matters more than the rate

At 0.3%, the base commission is small. A $40 order earns an agent developer 12 cents (our calculation). Nobody is building an agent business on that alone.

But the structure matters more than the rate. Three things stand out to us.

Paid and organic can sit in the same list. Shopify marks promoted products with a placement object, so a developer can tell them apart. Whether the shopper can tell them apart depends on how the agent displays them. Shopify requires labeling and disclosure, and it does not prescribe the wording.

Merchants can raise the rate. Through Shop Campaigns, merchants fund extra commission. That gives an agent developer a reason, at least in principle, to prefer products that pay more. Good developers will resist that. The incentive still exists.

The higher rate pulls purchases into Shop. The extra commission is paid only when the buyer finishes in the Shop app or on shop.com. Products with merchant-funded commission send the buyer to Shop, not the merchant's own store. For brands that sell direct on Shopify, that is a trade-off worth understanding before opting into a campaign.

There is one more detail worth noting for brands that are not on Shopify. Products from merchants outside Shopify can appear in the Global Catalog, but Shopify says they are not eligible for commission and never carry a placement marker. So an agent that leans on promoted placements could, in principle, favor Shopify merchants over a brand that sells through other platforms or only through retailers. We have no evidence that any agent does this. It is simply where the incentive points, and it is worth keeping in mind when you look at which products an agent built on Shopify's catalog tends to show.

Three companies, three positions

The interesting comparison is with the other large AI shopping surfaces, which have taken visibly different positions this month.

OpenAI's help page, updated this week, says ChatGPT's product results "are selected independently by ChatGPT and are not ads, nor influenced by any OpenAI partnerships." Ads in ChatGPT exist, but separately. Search Engine Watch reported on October 29 that OpenAI's new visual ad format had started showing inside chats in a limited US test, with an Ad label.

Google has been testing how AI Mode labels paid products. Search Engine Land reported on October 28 that Google was testing "Sponsored" and "All products" headings in AI Mode shopping results, which keeps paid and organic in separate, labeled groups. We covered the earlier ad changes in (post:ai-mode-ads-checklist-cpg-brands).

Shopify is doing something different. It is not an assistant itself. It supplies the catalog that many third-party agents will search, and it pays those agents. Each agent decides how to present what it gets.

Table: How three AI shopping surfaces handle paid products, October 2026

Company	Role	Stated approach to paid products
OpenAI	Runs ChatGPT	Product results are not ads; ads are separate and labeled
Google	Runs AI Mode	Testing labeled Sponsored and All products sections
Shopify	Supplies catalog to third-party agents	Pays agents commission on promoted placements; requires agents to disclose and label

None of these positions is wrong. But they mean that "an AI agent recommended it" will carry different meanings depending on which agent and which catalog.

The trust problem is the agent's, not the brand's

Most of the argument for AI shopping rests on trust. A shopper hands a task to an assistant because they believe it will find the best option for them. Amazon's Andy Jassy said this week, according to Fortune, that third-party shopping agents are "still not great, nor more convenient than coming to a retailer directly." Commission-driven ranking, if it shows, would be an easy way to confirm that skepticism.

That is why we think disclosure is the thing to watch, more than the rate. Shopify's rules say not to bury the disclosure in a footer and to make the paid relationship clear before buyers act. If agents follow that, shoppers can weigh a promoted result the way they weigh a sponsored listing on a retailer site. If agents do it badly, regulators and shoppers will notice, and the whole category will pay.

For brands, the risk is indirect. You do not control how a third-party agent labels your promoted product. If shoppers come to distrust agent recommendations in general, they will distrust your organic appearances too.

What we would do as a CPG brand

Keep organic and paid separate in your head and your reports. An appearance in ChatGPT's organic product list, a labeled ad, and a commission-backed placement in a third-party agent are three different things. Count them separately.

Win the organic list first. OpenAI says its merchant ranking weighs availability, price, quality and whether you are the maker, which we cover in (post:chatgpt-merchant-ranking-maker-primary-seller). That is earned with data, not money. Tip (tip:33) covers deciding whether to let agents buy from your store at all.

If you run Shop Campaigns, read the new terms. Merchants with external platforms turned on now include the Catalog API. Understand that commission-funded clicks land in Shop, and decide whether that is where you want the order.

Check how your products appear. Ask the agents your shoppers use about your category, and note whether results are labeled. Tip (tip:12) has a simple method; some teams use tools such as SearchPanel from XENA Intelligence to do it across many products.

Why affiliate money is not new

It helps to remember that commissions already shape a lot of shopping content. Review sites, deal newsletters and social creators have been paid on affiliate links for years, and the better ones disclose it clearly and still give honest recommendations. Shoppers have learned to read a "we may earn a commission" line and decide how much to trust the publisher. There is no reason AI agents cannot work the same way.

The difference is presentation. A review article has an author, a byline and a disclosure at the top. An agent's answer often arrives as a short list or a single suggestion in a chat window, with little room for context. If the agent picks one product for you, the question of whether that pick was paid matters more than it does on a page of ten options. That is why Shopify's rule that disclosure must come before the buyer acts, and must not be buried, is the most important line in its documentation.

Where this goes

Paid placement was always going to reach AI shopping. Search went the same way, and so did retail sites. The open question is whether agents will be clearer about it than search engines were in their early years.

Shopify has written the right rules for disclosure. OpenAI has drawn a firm line between its product results and its ads. Google is testing labels. If the industry keeps paid results visibly labeled, shoppers can keep trusting the organic part of the answer, and brands that earn it with good data keep their advantage. If labels blur, the most valuable thing about an AI recommendation, that it seems to work for the shopper, starts to wear away.