“Probably 99% of the issuer processing systems out in the world could process an agentic payment,” says Paymentology CTO Tim Joslyn. The issue is not the plumbing; the global infrastructure for moving money is already mature and functional. Instead, the challenge lies in recognizing these transactions and implementing the necessary identity and trust frameworks.
The real competitive battleground is not the payment rail, but the trust infrastructure that sits above it. As autonomous agents begin to execute transactions at machine speed, the industry must shift its focus toward orchestration, programmable controls, and the verification of intent.
The Architecture of Trust
The industry is currently engaged in a rapid build-out of these necessary trust layers. This is not merely theoretical; it is a response to the reality of deployments like Alipay’s, which has already rolled out delegated AI purchasing capabilities to over 100 million users. To support this scale, major players are establishing proprietary and interoperable frameworks to define how an agent proves its authority.
In April 2026, the FIDO Alliance launched an Agentic Authentication Technical Working Group, chaired by industry leaders including CVS Health, Google, and OpenAI, with vice-chairs from Amazon and Okta. Their objective is to create interoperable standards for AI agent authentication. Simultaneously, payment networks are codifying their own protocols. Visa has introduced its Trusted Agent Protocol (TAP), a cryptographic verification framework already seeing support from Adyen, Cloudflare, and Stripe. Mastercard is developing “Verifiable Intent,” a trust layer designed to provide provable user authorization for AI-driven transactions. American Express has launched ACE (Agentic Commerce Experiences), a developer framework that spans the entire lifecycle of an agentic transaction, from registration and account enablement to intent intelligence and cart context.
The Preparedness Gap
While the technical frameworks are under construction, the operational reality for merchants remains precarious. The Ravelin 2026 Agentic Commerce and Fraud Report highlights a significant disconnect between adoption and readiness. While 76% of enterprise merchants are currently integrating agentic protocols — with 44% already active and 32% planning to do so within six months — only 29% report feeling “very prepared” for the resulting fraud implications.
This gap is exacerbated by the fundamental shift in risk modeling. Joslyn points out that we are moving from human-speed fraud to machine-speed fraud. Traditional fraud systems rely on assumptions about human limitations, such as sleep cycles, behavioral consistency, and cognitive friction. Autonomous agents operate without these constraints, necessitating a complete overhaul of how risk is monitored and mitigated. Furthermore, while 53% of e-commerce merchants report that they trust AI agents more than human shoppers, the critical issue of liability remains unresolved.
The Consumer Trust Barrier
Beyond the technical and merchant-side challenges, there is a persistent consumer trust deficit. According to the Product.ai April 2026 Trust in AI Commerce Report, only 14% of consumers are willing to trust AI recommendations without manual verification. A significant 86% of users verify recommendations before completing a purchase, and 42% of consumers refuse to use AI for any transaction exceeding $25. This suggests that while the industry is building the infrastructure for autonomous purchasing, the end-user remains a significant bottleneck in the adoption curve.
The industry is moving past the speculative phase. As Joslyn observed, “Agentic commerce has stopped being a fantasy. It’s not a ‘Will this happen?’ It’s now a ‘How do we do it, and how do we scale it safely?’” The answer to that question will not be found in the payment rails themselves, which are already capable of handling the volume. Instead, it will be determined by the efficacy of the trust frameworks currently being built to manage identity, verify intent, and resolve liability in an environment where the shopper is no longer human.