AI’s New Age: Building Human Intent and Trust Into Agentic AI

Automation already drives most eCommerce today, but shopping is still personal. Agentic AI will gain acceptance only if it amplifies human choice, bringing personalization and ease where consumers want less friction.

Agentic AI is rapidly transforming commerce by simplifying friction points like returns and complex shopping tasks. Tomorrow, this intelligence will revolutionize loyalty programs, personalize negotiated offers and create comprehensive shopper profiles.

Agentic AI’s future depends on trust, not just tech. High levels of consumer and business wariness mean it is essential to implement interoperability now to enable consistent governance. Pre-authorization and strong security are vital for preventing fraud and scaling the ecosystem successfully.

Get Unlimit…

Get Unlimited Access Complete the form below for free, unlimited access to all our Data Studies, Trackers, and MonitorEdge reports.

Thank you for registering. Please confirm your email to view all our Trackers.

Agentic artificial intelligence (AI) is poised to redefine how money moves. The automation that currently powers reorders, renewals and recommendations is evolving into something more dynamic. Today, AI-powered agents can make decisions, negotiate and personalize on the shopper’s behalf.

However, these shopping proxies are still just that—proxies. They will not achieve their purpose by replacing human choice but rather by helping people fulfill and actualize their own. Consumers still want commerce to feel personal, expressive and secure. Therefore, agentic AI’s success will depend on trust, compliance and perfect alignment with human goals and behavior.

As agentic AI expands rapidly toward mainstream adoption, the real story isn’t about machines making choices. It’s about technology learning to understand and amplify the human intention driving those choices.

Aligning Agentic AI With Human Behavior
Today’s and Tomorrow’s Use Cases
Trust and Governance: The Keys to Agentic AI’s Future Success
Building the Future of Agentic AI With Trust, Humanity and Open Collaboration

Aligning Agentic AI With Human Behavior

Shoppers still seek identity-driven experiences, even as eCommerce becomes more automated.

Most eCommerce today is already automated—from subscriptions and reorders to bill payments and other low-engagement tasks. However, consumers are still shopping for items that showcase personal expression, such as apparel, home goods and gifts. For agentic commerce to fulfill its promise, AI must align with human behavior rather than work against it.1

Agentic commerce must be compatible with customers’ needs and expectations.

Insiders note that it is possible to train agentic AI to reflect customers’ preferences, creating a more seamless and satisfying eCommerce journey. This focus on individual desires and expectations mirrors lessons from past AI deployments, where personalization has proven essential to driving conversion. Eight in 10 consumers are more inclined to make purchases when brands provide a personalized experience. AI agents that offer personalization using key customer insights will be better set up for success.

Agentic AI’s rapid growth demands a sharper focus on the customer.

The technology is growing at a staggering rate. The agentic AI market is projected to increase from $7.06 billion this year to $93.2 billion in 2032. Analysts predict that 33% of enterprise software apps will use agentic AI by 2028. This figure marks a sharp increase from less than 1% in 2024. Amid this rapid expansion, it is essential for businesses to build in a focus on meeting customer needs.

One way that agentic AI can align with customer expectations is by reducing checkout friction. According to a new report from Spreedly, 29% of companies identify customer abandonment as the most common cause of failed checkouts, making it the leading culprit, even beyond unsupported payment methods. This indicates that consumers are having subpar experiences, perhaps due to friction or a lack of trust. Responsible agentic AI deployments can provide a more seamless purchasing journey.

Today’s and Tomorrow’s Use Cases

Agentic payments are already streamlining frustrating tasks and complex payments.

Agentic payments

are already simplifying friction-filled processes, handling complex shopping tasks and completing low-lift one-time payments.3

Experts note that the technology is currently delivering value for a number of specific use cases.2 Agentic payments are simplifying friction-filled processes such as returns and dispute resolutions. They are also handling complex shopping tasks like comparing products and assembling carts across multiple merchants. Additionally, agentic AI is completing low-lift one-time payments such as bills and taxes.

Looking ahead, firms will leverage agentic AI to boost loyalty and personalize offers.

Thought leaders envision a broader role for the technology in tomorrow’s use cases.4 Visa, for example, recently debuted its “Intelligent Commerce” application that enables AI agents to conduct transactions using tokenized credentials and real-time controls. With the shopper’s consent, the platform shares basic spend-pattern data to personalize offers and tailor recommendations. Meanwhile, Mastercard’s Agentic Tokens aim to provide more relevant shopper perks and targeted recommendations. Together, these innovations demonstrate how firms will use agentic AI to build loyalty and create personalized shopping experiences at scale.

Trust and Governance: The Keys to Agentic AI’s Future Success

Consumers and businesses alike are skeptical about entrusting AI to make payments.

At 56%, most consumers in the United States say they are not interested in using AI shopping assistants—and skepticism runs deep. More than four in 10 (41%) say they do not trust these assistants at all, while only 4% express complete trust in them. Another 34% cite concerns about privacy or security as key reasons for hesitation.

Businesses have similar apprehensions about AI payments. Spreedly reports that 22% of executives worry about losing control or transparency. An equal share express doubts related to compliance and false positives, where legitimate transactions are wrongly identified as fraud. Additionally, 20% are concerned that errors or lack of transparency in AI-driven payments could erode customer trust.

Trust and governance are key to unlocking the next phase of agentic adoption.

The building blocks for agentic AI payments already exist: tokenization, spend controls, delegated authority and real-time decisioning. What’s needed now is a framework of trust and governance to connect them. That means pre-authorizing agents based on intent and constraints, signaling trusted agent identity to payment networks and blocking bad actors before checkout.5

Industry momentum is shifting toward several standardized—and possibly competing—frameworks for agent authorization and verifiable delegation. As agent ecosystems scale, trust will depend on modular, interoperable governance built in from the start, not added later.

Consumer sentiment suggests the timing is right: Nearly one-third (32%) of Gen Z consumers say they’re comfortable letting AI agents purchase on their behalf. To succeed, agentic AI must empower choice and ensure that commerce feels not less but more human.

Building the Future of Agentic AI With Trust, Humanity and Open Collaboration

Agentic AI is reshaping how commerce operates, from automating complex transactions to anticipating consumer needs. As this technology matures, one principle is clear: Trust and human-centered design must guide its evolution. By embedding transparency, interoperability and user control into every stage of development, businesses can unlock agentic AI’s full potential while keeping customers firmly in charge.

PYMNTS Intelligence offers the following roadmap for organizations building or adopting agentic AI:

Build trust first. With 41% of consumers expressing distrust toward AI shopping assistants, brands must embed transparency, explainability and user control into every deployment.
Start where AI adds clear value. Early wins will come from low-risk, high-friction areas—such as renewals, returns and bill payments—where agents reduce hassle without replacing human choice.
Embed interoperable governance. Pre-authorization, agent identity verification and open standards will be essential to preventing fraud and ensuring ecosystem trust.
Collaborate across open platforms. Working within modular ecosystems, such as Spreedly’s, enables companies to build flexible, secure frameworks that scale responsibly.
Design with humanity in mind. When AI complements, rather than overrides, user intent, it builds lasting confidence, loyalty and long-term adoption.

Agentic AI is here—but its impact will depend on how it is configured. By prioritizing trust, governance and open collaboration, businesses can create payment experiences that are secure, human-centered and ready for the future.

Justin Benson

**The rise of agentic AI is more than a technological shift—it’s redefining trust and ownership within commerce flows. At Spreedly, we see this moment as an opportunity to pair speed and responsibility. Businesses are utilizing our platform for the rapid experimentation and deployment of their AI strategies, knowing they will be grounded in trust and transparency. Agentic commerce is unique and requires that innovation and controls advance together to deliver a more secure, seamless payment experience for everyone.”

Justin Benson CEO, Spreedly

1. Spreedly. The Future Is Agentic. PowerPoint presentation.↩ 2. Ibid.↩ 3. Ibid.↩ 4. Ibid.↩ 5. Ibid.↩