Retail marketers are starting to lose the moment of choice to AI as AI search and assistants are seeing increased use in shopping.

Commerce media and advertising technology company Koddi's survey of 750 consumers in the US, UK, and Germany found 64% of respondents find it appealing for an AI agent to "suggest brands I might like that I don't usually buy." That openness is already becoming behavior: We forecast nearly 80 million US consumers will use AI platforms and assistants for shopping this year.

As AI-assisted shopping becomes routine, trust is likely to follow: Koddi found that consumers who already use AI tools are far more confident in them, suggesting comfort grows with exposure.

Eric Brackmann, Koddi's vice president of commerce media, lays out the new reality: When an assistant filters choices and buys for the shopper, retail media shifts "from buying human attention to buying eligibility."

The report's own summary is sharper: "value is shifting at pace from those who are seen to those who are selected."

Shoppers are seeking a shopping co-pilot

Openness to suggestions is not permission to buy. In no category are more than one in seven consumers happy to let AI choose and act, and 72% want AI "to be a co-pilot, not a full autopilot," according to Koddi's data. The biggest trust breaker, chosen by 55%, is an AI making a purchase without approval.

One US shopper described the likely pattern in the report: "I think maybe after I've looked around and found a few options, I'd use AI to sort of narrow it down."

For retailers, the opportunity moves to the middle of the funnel. Brands need to earn a spot on the shortlist well before checkout.

Disclosure is the price of admission

Most shoppers accept paid placements in AI answers or overviews, but only in the open. About three-quarters want to know when a brand has paid to rank higher, yet 56% say paying for shortlist inclusion is fine or fine if clearly labeled, per Koddi. Paid placement still ranks among the top trust breakers, cited by 35%.

As shopping turns conversational, Brackmann cautions, "transparency can't be an afterthought."

Retailers that blur sponsored and organic results risk the trust erosion sponsored search suffered.

Retail media must move from surface to system

Retail media networks were built on owning the screen where shoppers decide. Third-party assistants effectively break that model, which Brackmann says "fractures" when the decision happens off-site.

Retailers still hold structural leverage: an assistant cannot confirm stock, local delivery, or regional pricing without retailer data. That advantage only counts if agents can read it.

Brackmann says retailers must keep data "connected, accurate, and sharp" or risk being "reduced to a commoditized fulfillment node."

Measurement is the gate to budget

Budgets are real but conditional.

• A third of advertisers plan to spend $50,000 to $249,000 on agentic commerce in the next year, and 61% would fund it from performance and paid search, Koddi found.
• Seven in ten want a 10% to 25% gain in ROAS or CPA before scaling.

Attribution is the obstacle. An AI recommendation can shape a purchase days later in another channel. Brackmann says the fix will take "more connected measurement frameworks, privacy-conscious data collaboration" across AI platforms, retailers and networks.

A US agency media director framed the dilemma in the report: "If there's real estate for only three brands to show up… how do I make sure, as a brand, to be on top?"

What retail marketers should consider

The biggest question for retail marketers is no longer how to get in front of shoppers, but how to get chosen by the AI that increasingly shops alongside them.

That likely means keeping product data clean and current enough for agents to read, putting more weight on mid-funnel consideration, labeling paid placements clearly, and working out measurement with AI platforms and retail partners before budgets scale.

Brand loyalty isn’t disappearing, but it will be earned inside the recommendation rather than on the shelf.

This was originally featured in the EMARKETER Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.