As artificial intelligence platforms and merchant-integrated agents become more sophisticated, consumers are expected to increasingly delegate tasks such as product searches, comparisons, decisionmaking and purchasing to agents. The shift signals a new era of commerce moving from direct user interaction to a model built on user authorization and autonomous agent execution.
Photo via China Daily
The rise of agentic commerce, however, presents new security and trust challenges. Merchants must be able to verify that transactions originate from legitimate AI agents rather than fraudulent actors, ensure purchases accurately reflect users' intentions and safeguard customer funds throughout the transaction process.
Against this backdrop, Chinese e-commerce giant JD unveiled its Agent Autonomous Payment Protocol on June 11. The framework aims to enable AI agents to complete payments independently within user-defined rules and spending limits.
The protocol establishes six levels of payment autonomy, providing a gradual pathway from transactions that require full human approval to those that can be executed entirely by AI agents. To enhance security, each payment request is linked to three critical elements: the verified identity of the user, the authorized version of the AI agent and a trusted operating environment. Transactions are approved only when all three conditions are validated.
The protocol also separates users' primary accounts from agent-driven spending. All agent transactions are routed through dedicated subaccounts subject to predefined controls, including spending caps, transaction limits, approved use cases, expiration periods and designated payees. Users can monitor activity in real time and revoke an agent's permission at any moment.
China's broader payment ecosystem is moving in the same direction. On May 26, Alipay debuted its AI Wallet and Token Pay, taking a step further to extend its next-generation AI payment infrastructure.
The initiative is also expanding beyond China.
Singapore-based Ant International introduced the Agentic Mobile Protocol, an agentic payment framework built for mobile interfaces including digital wallets, super apps and wearable devices. The protocol is designed to facilitate secure and seamless purchases through AI agents while helping merchants overcome challenges such as cross-border interoperability as agentic commerce scales globally.
A key component of the Agentic Mobile Protocol is an enhanced Know Your Agent authentication framework, which establishes a standardized trust-rating system for AI agents. The mechanism is intended to allow merchants to verify not only the identity of customers, but also the identity and trustworthiness of the agents acting on their behalf, helping ensure that transactions reflect genuine purchasing intent.
China's push into agentic payments comes as global payment networks race to build the infrastructure for agentic commerce. Industry leaders are increasingly developing technologies that will allow autonomous AI agents to search and complete transactions on behalf of consumers and businesses.
Mastercard is pursuing a vision of a future in which AI agents could eventually transact continuously with one another, executing chains of payments.
On June 10, the company unveiled Agent Pay for Machines, a new service designed to enable machine-to-machine transactions across its global payments network. The service will allow payments to be authorized, orchestrated and settled at machine speed, supporting emerging AI-driven business models.
Unlike traditional point-of-sale or person-to-merchant payments — which are initiated directly by users — machine-to-machine transactions are expected to be programmatic, continuous and executed between systems in the background of digital commerce.
"Machine payments can make it possible for services to be bought and sold among agents at fundamentally different scales than payments today — very high volumes, very small values, very fast and at extremely low latency," said Jorn Lambert, Mastercard's chief product officer.
Also on June 10, Visa announced a strategic collaboration with OpenAI to enable secure payments within agentic commerce. Under the partnership, Visa will provide its global payments network, credentialing capabilities and security infrastructure to support agentic commerce services. The company's payment capabilities will be integrated into OpenAI experiences, giving developers and merchants a streamlined way to accept Visa payments initiated by AI agents.
Visa said payments will operate within clearly defined user permissions and controls, including spending limits, approved merchant categories and required approvals. Transactions will be backed by tokenized credentials, real-time authorization and fraud-monitoring systems designed to maintain strong consumer protections.
"AI will transform commerce more profoundly than the internet or mobile technology ever did," said Jack Forestell, Visa's chief product and strategy officer. "As AI agents become active participants in the economy, Visa's focus is to ensure transactions are trusted, secure and seamless."