You are now leaving Synchrony.com. Please note that because the site you are going to is not controlled by Synchrony, you will be subject to the terms of use and privacy policy of the website you are visiting.
Michelle Romero: Michelle.Romero@syf.com Tyler Allen: Tyler.Allen@syf.com
Article
October 21, 2026, 4:39 PM EDT
Key Highlights
AI shopping is moving from helping consumers choose to acting on their behalf. That creates a new test for the companies building it: can an agent earn enough trust to make a recommendation, complete a purchase and remain accountable when something goes wrong?
Our 2026 AI in Commerce study, conducted with Oxford Economics, found that trust shapes both where consumers will begin using AI and how much authority they will eventually give it. That trust starts with the recommendation itself and extends to how an agent handles data and acts on the consumer’s behalf. Adoption is most likely to begin with low-risk tasks that deliver a visible benefit. From there, consumers will give agents more authority when the stakes are manageable, the provider is familiar and they remain in control. Accurate information, clear limits and reliable protections are needed at every step.
The Foundation: Trust in the Recommendation and the Transaction
According to the research, consumers begin judging an AI agent before it completes a purchase or handles their personal information. They need to know that the agent understood their request, weighed the factors they chose and looked beyond product and price. Security and data protection remain essential, but they cannot make up for a recommendation the consumer believes is wrong.
The research identifies three tests for trust.
Brands, retailers, financial institutions, payment providers and AI platforms need a common way to give agents accurate, complete and structured information about price, availability, service, protections, rewards and financing. Synchrony is contributing to open standards and testing agentic commerce through pilots with partners to help build and apply that shared infrastructure. The goal is to give agents a stronger basis for recommending an option that serves the consumer while clear controls and protections sustain confidence through the purchase.
Where Adoption Starts: Clear Value and Low Risk
The research also shows consumers are most comfortable with AI when it delivers clear value through low-risk tasks such as comparing options, saving money and applying discounts or rewards. Their comfort declines as AI moves from assisting with a decision to acting on their behalf:
The pattern is clear: consumers are most comfortable when AI delivers immediate, visible value. Their comfort declines as the agent moves from helping them make a decision to controlling the transaction.
For businesses, the adoption path starts with useful tasks that leave the consumer in control. An agent can take on more only after it has shown that it can make sound recommendations and act within the limits the consumer sets.
How Consumer Confidence Will Enable AI Shopping to Scale
The research also made clear that consumer confidence in AI shopping depends on three related factors: how much authority an agent is given, who provides the experience and what happens when the agent gets a decision wrong.
Higher-stakes purchases require more consumer control, stronger protections and a provider the consumer recognizes. A poor recommendation or incorrect purchase can quickly undo that confidence.
Building Trust from Recommendation Through Purchase
The research points to a clear sequence. Consumers first have to trust the recommendation, then the controls around the purchase and finally the protections available if something goes wrong. Each step depends on agents receiving accurate information and businesses remaining accountable for the experience.
Synchrony is working with partners to make financing, rewards, offers and protections understandable to AI agents and clear to customers. We’re also testing agentic commerce through pilots and working with industry and open-source groups on standards for accurate recommendations, customer approval and the handling of information and money. The next step is to turn those efforts into shared rules that work across platforms, so consumers can set the limits, agents can evaluate the full value of an option and every participant knows what it is responsible for when a purchase is made.
AI shopping is here. Whether it scales will depend on the trust consumers place in the recommendation, the control they retain over the transaction and the protections they can rely on if something goes wrong. Synchrony will keep working across the ecosystem to earn that trust, help brands and merchants make their full value visible to AI agents and turn consumer expectations into shared standards that work across platforms.
For more information, view the Synchrony and Oxford Economics 2026 AI in Commerce study and thought leadership piece.
Frequently Asked Questions
Q: What do consumers want from AI shopping and agentic commerce? A: Consumers want AI shopping to provide clear value while giving them security, transparency and control. The Synchrony and Oxford Economics research found that 82% of consumers want their data kept secure, 77% want clarity about how their data is used and 76% want to choose how much control they retain over AI.
Q: What makes consumers trust AI to make purchases on their behalf? A: Consumers are more comfortable giving AI authority when the stakes are manageable, the provider is familiar and they retain control over the transaction. The research found that 74% want to review and approve purchases an 67% would use AI more if fraud protection were included.
Q: How comfortable are consumers with AI making purchases? A: Consumer comfort varies significantly with the value of the purchase. For purchases under $50, 81% would allow AI some role in the transaction, including 47% who would let AI suggest options for approval and 34% who would allow AI to act on its own. For purchases of $5,000 or more, 46% would not use AI at all.
Q: What is the role of trust in agentic commerce? A: Trust helps determine how much authority consumers are willing to give AI agents as AI moves from recommendations toward transactions. The research shows that consumers evaluate the quality of an AI recommendation, how their data is handled, how much control they retain and what protections are available if something goes wrong.
Q: What happens when an AI shopping agent makes a mistake? A: Incorrect recommendations and purchases can cause consumers to stop using AI or even stop shopping with a retailer. The research found that 60% would stop or pause using AI after a poor recommendation, 74% would do so after an incorrect purchase and 44% would stop shopping with a retailer altogether after an unauthorized or wrong purchase.
Q: How is Synchrony preparing for agentic commerce? A: Synchrony is working with partners to make financing, rewards, offers and protections understandable to AI agents and clear to customers. Synchrony is also testing agentic commerce through pilots and working with industry and open-source groups on standards for accurate recommendations, customer approval and the handling of information and money.
Q: How could agentic AI change shopping and payments? A: Agentic AI can move beyond helping consumers evaluate options to acting on their behalf, including making purchases within parameters consumers set. The research shows that consumer willingness to delegate those decisions depends on clear value, accurate recommendations, controls and protections throughout the shopping and payment experience.
Q: What should merchants and brands consider as AI shopping grows? A: Merchants and brands need to make information about price, availability, service, protections, rewards and financing accurate, complete and understandable to AI agents. The research also indicates that businesses need to maintain clear consumer controls and protections as agents take a larger role in the transaction.