The $3–5 Trillion Problem
AI shopping agents are already here — systems that browse, compare prices, and complete purchases on your behalf with no clicking required. Projections cited by Ant International suggest that by 2030, AI agents will orchestrate between $3 trillion and $5 trillion in global consumer commerce annually. There's one enormous problem: nobody has a reliable way to verify that the agent making a purchase is who it claims to be, or that it's actually authorized by the person it's supposedly representing.
That gap between capability and trust is exactly what the new Know-Your-Agent (KYA) interoperability framework, announced on September 10, 2026, is designed to close.
Three Competing Protocols, One Framework
Until now, Visa, Mastercard, and Ant International each operated their own proprietary system for handling AI agent transactions. Visa built the Trusted Agent Protocol (TAP). Mastercard developed Verifiable Intent. Ant International created its Agentic Mobile Protocol (AMP). These systems couldn't talk to each other, meaning every merchant, wallet, and agent platform that wanted to operate across networks had to integrate each protocol separately — a costly, duplicative mess.
The KYA interoperability framework aligns these three protocols without replacing them. Think of it less as a new standard and more as a shared translation layer — a common set of principles and signals that allow each network's own verification system to recognize trusted agents from the others.
Three Pillars of the Framework
The framework rests on three core mechanisms:
- Cross-network operator traceability: Links an AI agent to a validated operator — a human, company, or cardholder — so every transaction has a traceable human principal behind it. The agent shops; the human is accountable.
- Shared certification requirements: Agents are assessed against agreed behavioral and security standards before being granted trusted status across networks. It functions like card-terminal certification, applied to software.
- Continuous transaction monitoring: Identity signals and payment behavior are monitored in real time, combining agent identity with transaction data for dynamic risk assessment even after an agent has been certified.
Why This Matters
As Unite.AI explains, the framework aims to eliminate duplicative agent verification, reduce integration complexity, and lower time-to-market for agentic commerce services. In plain terms: a merchant or payment network can now confirm that an AI agent making a purchase is genuinely authorized to do so — across Visa, Mastercard, and Ant's ecosystems simultaneously.
The precedent-setting aspect matters as much as the technical details. Visa, Mastercard, and Ant International together represent a dominant share of global payment infrastructure. Their alignment on shared agent identity principles will likely become the default baseline that other payment networks follow.
The hard part, as Forkast noted, is now ahead: aligning three global networks on principles is one thing; building the real-time cross-network verification infrastructure to run it at scale across millions of merchants is another. But the KYA framework's existence means the industry has at least agreed on what problem it's solving — and that's the prerequisite for solving it.