FIS Says Agentic Commerce Needs Proof, Not Promises

The shift from transactional to relational payments is accelerating. We are entering the era of agentic commerce, where AI agents act on behalf of shoppers to source, negotiate and complete purchases. Artificial intelligence is no longer just a back-end tool — it is becoming the primary engine of customer engagement, which represents a new, unique opportunity for brands to reimagine consumer lifecycle engagement strategies that have been in place for many years. That shift has forced us to rethink governance frameworks we had taken for granted.

With AI agents projected to help orchestrate up to $1 trillion in U.S. retail revenue by 2030, AI algorithms are already influencing real-time decisions, financial access and customer trust. The core question is no longer whether we adopt AI, but how we govern it to ensure shared success across the ecosystem.

Where Governance Breaks Down

In our experience, AI governance most often breaks down at the point of integration. We see ecosystems where AI models operate in silos — where purchase-event signals and item-level intelligence do not communicate securely within a single infrastructure. When that happens, visibility is lost. Governance fails when organizations treat AI solely as a post-purchase optimization tool rather than embedding it directly into authorization, authentication and dispute networks from the start. We learned early that governance has to be architected into the payment flow itself, not layered on after the fact.

The Speed vs Governance Trade-Off

The hardest trade-off we navigate is balancing the rapid deployment of hyper-personalized offers against the strict governance of item-level data. Consumers expect frictionless, instant rewards at checkout. But building solutions we can stand behind has required us to prioritize secure, receipt-backed proof of performance over simply rushing a model to market. Speed without precision creates risk at scale. That has meant making deliberate choices to slow down in places where others might not — and those choices have paid off in the durability of what we have built.

Governing Across Third Parties

Governing AI across third-party networks requires a funder-agnostic approach grounded in universal standards. Card issuers, merchants and consumer brands must operate within a tightly integrated ecosystem where intelligence travels securely with the payment itself.  Whether we are analyzing the purchase event or the specific items in a basket, we enforce strict data guardrails so that agents can transact safely across all participating partners. That discipline is what makes a four-party model — consumers, merchants, issuers and brands — actually work.

The Decision I Wish We Had Made Earlier

If there is one governance decision I wish the industry had made sooner, it is defining exactly what AI can “see” and act upon. Establishing firm boundaries around item-aware decisioning from day one — clearly distinguishing between a total-receipt view and a digital-receipt view, for example — prevents privacy bottlenecks downstream and builds consumer trust that agentic commerce depends on. We have seen firsthand how retroactively drawing those lines is far more costly than setting them at the outset.

The Question Boards Should Be Asking

Executive teams frequently ask whether their AI infrastructure is secure. The better question is: how does our AI governance scale when algorithms, rather than humans, initiate the purchase?

In an agentic world, trust is the ultimate currency. The data is available and the technology is ready. The leaders who invest now in governance frameworks that turn every AI-initiated interaction into a secure moment of value creation will define the next era of payments.