VeChain Explains Why Blockchain Could Become the Trust Layer for AI Agents
As artificial intelligence tools become more widely available, the competition between AI agents may increasingly depend on factors beyond the underlying models. Sunny Lu, associated with VeChain, has outlined a five-layer framework for understanding the AI ecosystem, arguing that blockchain could play a key role in establishing trust between agents, users, and digital services.
In a post, Lu suggested that the differentiating factor for AI agents may have less to do with the model itself than many people expect. His framework spans the infrastructure powering AI, computing hardware, models and data brokers, agents, and user interfaces. Within this structure, he places blockchain in the trust layer supporting agents.
The argument reflects a broader question facing the AI industry: as access to capable models expands, what will make one agent more useful or reliable than another? Performance remains important, but the ability to interact securely with other systems and establish trust could also influence adoption.
VeChain Executive Breaks Down the Five AI Layers
At the foundation of Lu’s framework is infrastructure, including power grids, internet data centers, and cloud services. These systems provide the physical and digital resources required to run AI workloads. Without sufficient electricity, data center capacity, and cloud infrastructure, more advanced applications cannot operate at scale.
The second layer consists of GPUs and computing power. AI models require substantial processing resources for training and inference, making hardware availability and computing efficiency important factors in the development of AI services. This layer connects the physical infrastructure supporting data centers with the software capabilities delivered to users.
The third layer includes models and data brokers. Models provide the capabilities behind AI systems, while data brokers represent another part of the information ecosystem that can support their operation. As model capabilities become more accessible, competition could shift toward how effectively developers combine these resources with specialized tools and reliable information.
The fourth layer is where Lu places AI agents, dividing it into capabilities and a trust layer. Capabilities include the tools an agent can use to perform tasks, while blockchain is presented as a potential foundation for trust. Under this view, an agent’s value could depend not only on its ability to complete a task but also on how reliably its interactions and transactions can be verified.
The fifth layer consists of interfaces, including robotics. These are the points through which people or other systems interact with AI capabilities. Interfaces can take the form of software applications or physical machines, connecting underlying AI systems to practical tasks and everyday use.
Why Blockchain Could Matter as AI Agents Expand
Lu’s broader argument is that AI will change more than productivity. He also expects changes in product relationships, with blockchain playing a role in completing that transformation. The post does not specify a detailed implementation, but the framework points toward a future in which AI agents interact with digital services and other agents through systems that need reliable ways to establish trust.
Blockchain networks can provide shared records and verifiable transaction histories, depending on how they are designed and implemented. These properties could be useful when agents exchange value, coordinate actions, or interact across organizational boundaries. However, blockchain alone does not guarantee that an AI agent will behave correctly, protect private information, or make sound decisions.
The practical impact will depend on how developers integrate blockchain with agent capabilities, identity systems, permissions, and security controls. Different applications will also have different requirements, meaning not every AI agent will necessarily need blockchain infrastructure.
For VeChain and the wider blockchain sector, the opportunity outlined by Lu is to become part of the infrastructure supporting interactions in an increasingly automated economy. As AI models become more widely available, trust, coordination, and integration could become more important areas of competition. Whether blockchain becomes a central part of that shift will depend on the real-world systems developers build and the problems those systems solve.
About The Author
