The transition from AI as a conversational tool to AI as an autonomous actor requires more than just smarter models; it requires a way to prove who is acting and what they are allowed to do. To address this, Zetrix AI Berhad and the China Academy of Information and Communications Technology (CAICT) have introduced a blockchain-AI trust layer for agentic AI ecosystem development, designed to provide a verifiable identity framework for autonomous software agents.
Unveiled on April 14, 2026, during the World Internet Conference Asia Pacific in Hong Kong, the initiative centers on a new platform called Avatar. The system is positioned as critical infrastructure for an emerging “agentic economy,” a shift where AI systems do not merely suggest actions but actively execute transactions, negotiate contracts, and communicate on behalf of individuals, and corporations.
By merging blockchain-based verification with AI autonomy, the platform aims to solve the persistent issues of impersonation, permission errors, and “hallucinations” that plague open AI environments. The goal is to move beyond simple chatbots toward a world of “digital twins”—agents trained on a user’s specific personality, knowledge, and communication style that can operate independently while remaining tethered to a verified human or corporate identity.
The Architecture of Trust: How Avatar Works
At its core, Avatar is not a standalone application but a trust protocol. It is built using a hybrid OpenClaw and pipeline framework, integrating the Zetrix Layer-1 public blockchain and the Astron blockchain. This technical foundation allows AI agents to access “verifiable credentials,” which include professional qualifications, legal identities, and digital assets.
This integration solves a primary friction point in agentic AI: the “authorization gap.” When an AI agent attempts to execute a payment or access a secure database, the receiving system needs a guarantee that the agent is legitimate and authorized by the human owner. Avatar provides this by acting as a settlement and identity layer, ensuring that the credentials used are real and the permissions are current.
To further encourage institutional adoption, the platform incorporates a security framework certified by CAICT, a prominent scientific research institute under China’s Ministry of Industry and Information Technology. The platform includes an open agent task store, allowing third-party developers to create and publish specialized agents that can be integrated with user credentials.
“Blockchain-empowered agentic AI will form the trust foundation for next-generation autonomous digital interactions,” said Mr. You Xiao Yu, Vice President of Astron CAICT. “As agents begin acting on behalf of individuals and enterprises, the secure and trustworthy execution becomes non-negotiable.”
Defining the ‘Agentic Economy’
The concept of an agentic economy suggests a future where the primary unit of internet interaction is no longer a human clicking a link, but an agent negotiating with another agent. This “Agent-to-Agent” (A2A) interaction model allows for seamless collaboration and transaction without immediate human intervention.
During the Hong Kong showcase, CZ Wong, Chief AI Officer of Zetrix AI and the architect of Avatar, detailed several specific use cases for this technology:
- Corporate Efficiency: Executives can deploy avatars to onboard new employees, answer strategic questions, and preserve institutional knowledge, effectively extending their availability.
- Scale for Public Figures: Celebrities and influencers can use avatars to manage fan engagement and commercial interactions, enabling thousands of personalized exchanges simultaneously.
- AI-Native Monetization: Creators can package their expertise into agents that run paid communities or deliver personalized teaching, creating a new revenue stream beyond their own real-time bandwidth.
- Autonomous Outreach: Companies may shift from static websites to corporate agents that actively seek prospects and tailor messaging to individual customer profiles.
- Digital Legacy: For consumers, the platform envisions a “super intelligent extension” of the self that could eventually serve as a verifiable digital legacy for future generations.
Strategic Implications and Market Impact
The partnership between Zetrix AI—headquartered in Malaysia—and CAICT signals a broader effort to standardize how AI identity is managed across borders, particularly between ASEAN and China. By utilizing Blockchain-based Identifiers (BID) and Verifiable Credentials (VC), the system attempts to create a gateway for global trade that is both automated and secure.
TS Wong, Group Managing Director of Zetrix AI, noted that agents representing human users herald a wave of tasks that strictly require knowing who is transacting and verifying the assets required to complete the task. This positions the blockchain-AI trust layer for agentic AI ecosystem as a fundamental shift in how we perceive the “internet of agents.”
| Feature | Function | Technical Driver |
|---|---|---|
| Digital Twins | Personality/Knowledge Replication | Agentic AI Training |
| Identity Layer | Verification of Representation | Zetrix/Astron Blockchains |
| Agent Store | Third-party Specialized Agents | Open API Framework |
| Security Framework | Enterprise-grade Trust | CAICT Certification |
The overarching bet is that trusted agent infrastructure will be as pivotal to the next era of the internet as websites were to the first and mobile apps were to the second. If successful, the platform would move the AI industry away from “isolated” assistants toward a networked ecosystem of verified autonomous actors.
The next phase of the rollout involves the integration of third-party developers into the agent task store and the expansion of CAICT security certifications for enterprise-grade agents. Official updates regarding developer access and the broader rollout of the Avatar platform are expected to be shared via the official Avatar portal.
We would love to hear your thoughts on the rise of the agentic economy. Do you trust an AI “digital twin” to handle your transactions? Share your views in the comments below.
