Merchants See a Bigger AI Prize if Consumers Trust the System
Merchants See a Bigger AI Prize if Consumers Trust the System
Publish Date: 2026-04-23 04:03:00
Source Domain: www.pymnts.com
Autonomous commerce is no longer a future state. It is already reshaping how consumers spend. Scaling it, new data shows, demands more than capable technology. It demands trust.
Findings in the new PYMNTS Intelligence report, “Embedded Offers: The Billion-Dollar Opportunity Inside Recent Consumer Spending,” produced in collaboration with FIS, find that adoption of autonomous and agentic artificial intelligence commerce does not follow from exposure or familiarity alone. It depends on permission design and specifically whether the system aligns with what users expect in terms of control.
The stakes are concrete: Building for automation capability without addressing expectations around trust, transparency and responsible control produces technically sophisticated products that fail to achieve broad uptake.
The data shows a majority of consumers are open to some level of automated offer management, but individual willingness varies significantly depending on the degree of control they relinquish.
Consumers are most comfortable with automation in lower-stakes contexts, such as selecting the best payment method. They are far less comfortable when systems make decisions about product choice, spending limits or brand substitution.
A Market Divided by Trust
The report data points to a structural division within the consumer base. Roughly half of consumers are receptive to some form of automated offer management, while a substantial minority remain cautious or resistant. On one side are the “automation-receptive” consumers who are willing to delegate decision-making in exchange for convenience and savings. On the other are “control keepers,” who can require explicit consent for every action and are frequently reluctant to share their data.
Advertisement: Scroll to Continue
A system calibrated for one group may alienate the other. In this sense, the challenge is not technological. The capability to automate offer selection and application already exists….
