Thursday, September 3, 2026

Sui Builds Infrastructure for Agentic Commerce

Neon-lit AI agent coordinates atomic payments on a Sui-like blockchain hub, with holographic mandates and AP2 icons.

The Sui Foundation is positioning its blockchain as infrastructure for agentic commerce, a model in which autonomous AI agents move beyond generating recommendations and begin executing financial tasks directly. The core challenge is giving those agents verifiable authorization, shared transaction state and reliable execution without requiring constant human intervention.

In an official Sui blog post on blockchain infrastructure for AI agents, the Foundation highlighted Sui’s role as a launch partner for the Agentic Payments Protocol, or AP2, an open standard co-developed by Google and Mysten Labs. AP2 is designed to let autonomous agents prove user intent and payment authority across different merchants and payment systems.

AP2 Adds Verifiable Permissions for AI Agents

A central component of the framework is the use of “mandates,” cryptographically signed digital contracts that define what an agent is authorized to do on behalf of a user. These mandates establish explicit limits around agent actions, creating a verifiable record of the permissions behind a transaction rather than relying solely on an application’s internal logic.

AP2 also sits alongside other emerging agent standards. The framework builds on Agent2Agent for communication between autonomous systems and the Model Context Protocol for handling AI context. The objective is to make agents interoperable across different platforms while preserving clear authorization and execution rules.

Payment authentication is another part of the emerging infrastructure. Organizations including Visa and Fiserv are involved in developing the Trusted Agent Protocol, which focuses on identifying legitimate agent-initiated transactions and reducing fraud risk. Agentic commerce requires payment systems to verify not only the transaction but also the authority of the machine initiating it.

Sui’s architecture is intended to support complex workflows through atomic execution. Multi-step operations can be processed as a single transaction, meaning they either complete fully or fail without leaving intermediate actions unresolved. Atomic execution reduces the reconciliation problems that could arise when autonomous agents coordinate several financial actions at once.

Sui Targets Real-Time Autonomous Payments

The model is already being tested through practical integrations. Skyfire recently demonstrated an agentic commerce transaction using its KYAPay protocol alongside Visa’s Intelligent Commerce suite. These systems aim to close the gap between an AI agent deciding what action to take and actually completing a payment through stablecoins or traditional financial rails.

Sui is also highlighting technical features such as sub-400 millisecond transaction finality and zkLogin as infrastructure for real-time programmable payments. Fast settlement and authentication are being positioned as key requirements for autonomous financial activity, where agents may need to execute transactions continuously and without manual approval at every step.

Potential applications range from automated operating expenses for small businesses to DeFi agents managing yield strategies. The broader vision is for AI systems to become transaction-capable participants rather than purely informational tools, operating within explicit permissions and auditable on-chain rules.

The Sui Foundation has also pointed to further protocol development, including confidential on-chain payments with controlled audit access targeted for 2026. The success of agentic commerce will ultimately depend on whether autonomy can be combined with privacy, compliance and enforceable user control at scale.

Scroll to Top
Chain Report
Privacy Overview

This website uses cookies so that we can provide you with the best user experience possible. Cookie information is stored in your browser and performs functions such as recognising you when you return to our website and helping our team to understand which sections of the website you find most interesting and useful.