Key Points

  • Zetrix AI Berhad and China’s CAICT launched a Know Your Agent (KYA) service on 5 November 2026, announced at the World Internet Conference in Wuzhen and the ICT Development Forum in Dubai.
  • The service gives AI agents verifiable digital identities linked to a business or a person, so others can check who an agent acts for and what it is allowed to do.
  • Owners can set limits on an agent, such as which services it may use, who it may pay, and how much it may spend, with human approval required above a set amount.
  • The system builds on W3C Verifiable Credentials and uses AvatarLabs’ Agentic Wallet technology to manage credentials and keep records of agent activity.
  • A cross-border pilot will run with organisations in the Middle East, Africa, China and ASEAN, and the findings will be submitted to the ITU, IEEE and other standards bodies.

Zetrix AI and China’s CAICT launched a Know Your Agent service on 5 November, giving AI agents a verifiable digital identity so businesses and governments can check who an agent works for before letting it act.

What happened

The announcement came on the same day from two stages: the World Internet Conference in Wuzhen, China, and the ICT Development Forum 2026 in Dubai, United Arab Emirates. Zetrix AI Berhad, a Malaysian company headquartered in Malaysia, developed the service with the China Academy of Information and Communications Technology (CAICT).

The idea is simple, and it answers a question that has grown urgent as AI agents take on real work. These agents now book services, move money, read business data and talk to other agents with little or no human supervision. Until now, there has been no standard way to prove that an agent is who it claims to be, or that it is allowed to do what it is trying to do. The KYA service is designed to close that gap.

Each participating agent gets a verifiable digital identity tied to an authenticated person or organisation. Owners then set clear permissions: which services the agent may reach, what information it may use, who it may transact with, and how much it may spend. A business could, for example, let an agent pay approved suppliers up to a set limit, while any payment above that limit needs a human to sign off. An agent seeking access to a government or company system would first have to show its identity and its authorisation.

The controls travel with the work. When one agent hands a task to another, the second agent cannot gain powers beyond those the first one held. Permissions can be taken away, and signed records of what each agent did create a trail that can be checked later.

“AI is entering a new era where intelligent agents will increasingly make decisions, interact with businesses and execute transactions independently. Greater autonomy must come with greater accountability. The question is no longer just what AI can do, but who is responsible for what it does,” said TS Wong, Group Managing Director of Zetrix AI, at the launch.

Under the surface, the service builds on W3C Verifiable Credentials, an open standard for digital proof, and uses AvatarLabs’ Agentic Wallet technology to hold credentials, check delegated permissions and keep evidence of agent activity.

Why it matters

Trust is now the bottleneck for AI agents, not capability. A model that can act is only useful if the people on the other side of the transaction believe it is genuine, authorised and accountable. Without that, every agent looks the same to a bank, a supplier or a government portal — and the safe answer is to refuse them all.

This launch matters because it comes from a Malaysian company working with a Chinese state research institute, and because it is aimed squarely at cross-border use. Agents do not stop at national borders, but identity systems do. A pilot spanning the Middle East, Africa, China and ASEAN will test whether an agent’s identity and permissions can be recognised in one country when they were issued in another, while each country keeps control of its own identity rules and trust decisions.

The same week underlined how fast AI is spreading through the region. Tencent’s Hy4 brought visual understanding and computer control to one of China’s biggest open model families, while AWS’s Semarak AI Malaysia programme put AI tools into the hands of government officers and small firms. Identity and accountability are the missing piece that lets that wave of agents operate safely in trade, payments and public services.

“As Agentic AI assumes greater responsibilities, the need for KYA services will continue to rise. It is our expectation that this joint service with Zetrix AI will help foster a cross-border trusted network enabling interoperable Agentic AI around the world,” said Xiaoyu You, Head of International Department, ASTRON network, CAICT.

What happens next

Three things to watch. First, the pilot: which organisations join, and whether the system works in live cross-border trade, digital payments and government services. The pilot’s participants are still subject to confirmation, and no general availability date or pricing has been announced. Second, the standards track: Zetrix AI and CAICT plan to submit the pilot’s findings to the International Telecommunication Union’s Study Group 17, through its Focus Group on Trust and Identity for Humans and Agentic AI, as well as to the IEEE and other standards bodies. Third, the wider race: other identity providers and governments are working on the same problem, and the approach that wins adoption first may shape how agents are identified everywhere.

Sources: PR Newswire / Zetrix AI Berhad, 5 November 2026 (via Morningstar); KLSE Screener, 5 November 2026. Quotes are from the launch announcement. Pilot participants are subject to confirmation; no pricing or general availability date has been announced. Published: 5 November 2026, Kuala Lumpur time.