Ant Group Launches APASS, a Trust Infrastructure for Agent Commerce

By AI123

At the AI Payment Forum of the 2026 Inclusion·Bund Summit held on September 11, Ant Group unveiled APASS, a trust infrastructure designed for commercial activities involving AI agents. Built on the Know Your Agent (KYA) concept, APASS establishes two trust chains covering identity and behavior, offering identity registration, continuous verification, intent security and trusted attestation. According to the company, APASS has been integrated with more than 1 million agents, with 110,000 activated, and covers 442 service providers supporting agent types including independent assistant, productivity, hardware terminal and merchant service.

As agents take on tasks such as searching, shopping, ordering and even corporate procurement, the traditional trust model centered on "person-account-platform" is no longer sufficient. APASS verifies an agent's identity and authorization before a transaction, detects abnormal behavior and intent deviation during the transaction, and retains evidence after the transaction for audit and accountability. On the identity side, it uses a three-layer model of static identity, runtime identity and responsible entity identity, combined with a six-dimensional continuous verification mechanism, to provide ongoing confirmation of an agent's identity, authorization, intent and behavior.

Chen Shupeng, head of AI payment security at Ant Group, said agents are becoming a "new population" in the business world, and the industry must answer not only "who is trading" but also "why they trade." Based on trusted identities, platforms can gradually open services such as search, service invocation, business processing and payment to agents according to risk levels.

More than 50 organizations, including Alibaba, Li Auto, Xiaomi, OPPO, Zhipu and the China Academy of Information and Communications Technology (CAICT), have joined the IIFAA intelligent agent trusted ecosystem initiative. Meanwhile, Alipay+ AMP, Mastercard's verifiable intent and Visa's intelligent agent trust protocol are expected to form a KYA interconnection framework, supporting agent identity verification and trust collaboration in cross-border scenarios.

Explore topicsAgentAnt GroupBund ConferenceModels

Sources

View all