AI Turns Scattered Payment Data Into Smarter Routing Decisions
Watch more: What’s Next in Payments With Spreedly’s Adam Hiatt Artificial intelligence’s near-term payments value may come less from autonomous checkout than from making the payment stack itself more intelligent and machine-readable. “There’s a ton of payments optimization potential in these payments rails, particularly in complex multi-PSP [payment service provider] environments,” Adam Hiatt,…
Artificial intelligence could revolutionize the payments industry by turning scattered payment data into smarter routing decisions, according to Adam Hiatt, executive vice president of product strategy at Spreedly. While AI agents may not initially focus on autonomous checkout, they can optimize complex payment rails, particularly in multi-PSP environments.
The technology can transform payments systems' overwhelming data into actionable insights, providing cost savings and efficiencies. AI's potential lies in its ability to make payments intelligence an enterprise decision layer, rather than a specialist function. By analyzing data from multiple payment service providers, processors, geographies, and payment methods, AI can continuously optimize complexity, rather than just periodically.
However, the near-term opportunity for agentic commerce may not be evenly distributed across categories. The first winners may be merchants whose inventory can be easily found, interpreted, and surfaced within AI-native discovery environments. Meanwhile, merchants should prioritize fixing their product catalog to optimize for AI-driven product discovery.
AI agents moving from recommendation to execution introduce new challenges, such as determining the legitimacy of machines and ensuring that authorized actions reflect the consumer's intent. This new challenge, known as the "know your agents" (KYA) model, is a significant development in agentic commerce that goes beyond traditional fraud detection.
Written by urgent.news from PYMNTS's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
