The Maze: Agentic commerce has an awkward math problem. The most monetizable user is someone who accepts AI ads and lets an agent buy for them. In a January 2026 Mod7 Research Strategy survey, that group was only 4% of US genAI assistant users. The biggest group, at 46%, rejected both. The future may still arrive. It is just not yet permissioned.
The near-term addressable market is smaller than the pitch deck. The EMARKETER data splits US genAI assistant users into four groups: 15% only open to ads in assistants, 36% only open to agentic commerce, 4% open to both, and 46% resistant to both. That is not a rejection of AI shopping as a concept. It is a rejection of bundling persuasion and delegation too quickly. The customer may accept help finding a product before they accept sponsored answers. They may accept an agent placing a repeat order before they let it choose a brand. The commercial path is likely modular, not magical.
Commerce utility beats ad acceptance, but only with a human-shaped guardrail. Millennials and Gen X are the most receptive to agentic commerce only, both at 43%. Gen Z is at 31%. Boomers fall to 19%. The support evidence points in the same direction: 58% of consumers say they are open to placing orders through an AI assistant, but only 6% have actually done so. That gap matters. It says the promise is legible, but the behavior is not normal yet. Retailers should start with low-risk journeys: replenishment, saved preferences, price tracking, post-purchase support, and cart preparation. Let the agent do the background work. Keep the buyer in the moments where trust, identity, and regret risk still matter.
The ad model carries the bigger trust tax. Across all users, 36% are only open to agentic commerce, while 15% are only open to AI ads. That difference is the strategic clue. People can imagine AI saving time. They are less excited about AI selling them something while it claims to help. A separate EMARKETER piece says 63% of US adults believe ads in AI search results would make them trust those results less. That is expensive friction if AI-search ads are priced like premium media. Retailers can test the channel, but they should not confuse early inventory with durable consumer permission.
Younger users are not a free pass. Gen Z has 49% resistant to both formats, nearly the same as Boomers at 50%. The familiar digital-native shortcut does not work here. This is not just an age adoption curve. It is a control problem. Users may like AI for search, summaries, or recommendations, then still distrust the moment where the assistant becomes a salesperson or a buyer. That makes transparency part of the product: why this answer, who paid, what data was used, whether the recommendation is sponsored, and where the final purchase authority sits.
Why it matters: Agentic commerce will not scale because retailers add checkout buttons to assistants. It will scale when users believe the agent is working for them, not arbitraging them. The first winners will separate assistance from monetization, prove savings or convenience, and ask for more autonomy one step at a time. The loser move is to bolt ads onto the assistant and call it a shopping revolution.
