Stablecoins to Be ‘Core’ to AI-to-AI Commerce, EastPoint: Seoul 2026 Panel Says

Surge in micropayments for API and data purchases

Card networks and blockchain payment rails may converge

Stablecoins are set to emerge as a core payment tool as "agentic commerce" spreads, with AI agents finding, purchasing and paying for goods and services on behalf of humans, panelists said. Handling repeated micropayments made in rapid succession will require payment rails that are programmable and can settle around the clock, they said.

A panel discussion at EastPoint: Seoul 2026 on Sept. 28 at the Westin Seoul Parnas in Seoul's Samseong-dong focused on payment infrastructure for the AI agent economy. The session was moderated by professor Lee Jong-seop and featured Joe Hwang, regional lead at Google Cloud; Cindy Xu, chief marketing officer at Kite AI; Jensen Tung, co-founder of Virtual Protocol; and Dennis Won, chief technology officer at Newton Labs.

Hwang said AI agents could advance to the point where they fully replace human judgment in transactions. Still, he expects the shift to be gradual rather than an immediate replacement of existing payment systems, moving from a structure in which humans make decisions to one in which agents transact autonomously.

"One of the clearest new areas to emerge as AI agents spread is micropayments," Hwang said. "A new economy could take shape in which countless small transactions occur over short periods, such as API calls and database searches."

Panelists also said consumer payments and transactions between AI agents may rely on different payment methods. Xu said end users in e-commerce will still feel more comfortable paying with cards and traditional currencies, while stablecoins are better suited to areas where frequent machine-to-machine micropayments take place, such as API fees and data purchases.

She added that card networks and stablecoin rails are likely to converge over time. Users may continue to see familiar payment methods at the front end, while settlement between agents in the background is handled through blockchains and stablecoins. Use cases are separate for now, but stablecoin adoption will broaden as agent-to-agent transactions increase and the need for faster settlement grows.

Tung said complex commerce between AI agents will require more than a simple payment network. Commercial activity also involves contract terms, service fulfillment and dispute resolution.

He said machine-to-machine commerce could be handled more efficiently if two AI agents agree on transaction terms through smart contracts and use them to escrow and settle funds. In that setup, smart contracts can automate verification of transaction terms and payment while reducing the risk of disputes between agents.

Xu also said giving agents autonomy must be paired with tightly defined limits on their authority. She compared delegating work to an AI agent to handing a child a credit card. Rather than simply passing over one card, users need to assign different permissions depending on the task, country, type of transaction and spending limit.

"To let agents process millisecond-level transactions autonomously, you need appropriate delegation of authority," Xu said. "Kite AI applies a structure that sets permissions separately for humans, agents and individual task sessions."

Won said blockchains are strong in fast settlement but still fall short when it comes to approving and controlling agent transactions. "Stablecoins are money that operates through software, but using them safely requires an approval framework that sets transaction limits and conditions," he said. "You need to be able to verify an agent's identity and check regulatory compliance."

Accountability was also cited as a key challenge. It remains unclear who should bear responsibility when an AI agent executes a transaction outside a user's instructions: the user, the AI model developer or the platform. Won said an agent's identity is the starting point for determining responsibility when problems arise, and standards are needed to verify the agent itself, beyond know-your-customer procedures designed for humans.

Hwang said the accountability structure used in traditional finance can also be applied to agentic commerce. If there is a problem with delivery of a product or service, the seller could bear responsibility. If an issue arises within the scope set by the user, the user could be responsible.

"Google's role is to provide the infrastructure and tools that allow AI agents to be developed and deployed," Hwang said. "Given the possibility that AI models may produce incorrect outcomes, deterministic controls such as identity verification, security and spending limits need to be built alongside them." He added that if an agent acts outside the authority it was given, responsibility on the part of the platform or service provider should be reviewed.

Kim Soo-hyun, Hankyung reporter / Hwang Doo-hyun, Bloomingbit reporter ksoohyun@hankyung.com