Pantera Partner: The rapid popularity of AI agents benefits blockchain, but most investors underestimate the opportunity Author: Cosmo Jiang&Sam Lehman Compiled: Deep Tide TechFlow Deep Tide Introduction: Pantera Capital partners present a clear argument in this article: AI agents do not require blockchain in the first few stages of automation, but once they enter the stage of fully autonomous trading between agents, the traditional financial track will completely fail. This is not a general narrative, but rather provides specific evidence from three dimensions: identity, payment, and trust. For readers who want to understand why the narrative line of "AI+encryption" holds true, this article is worth reading carefully. The viral rise of Cosmo Jiang, Sam LehmanOpenClaw (formerly known as Clawdbot) marks a generational leap in autonomy. When these AI agents start interacting with each other - in some cases negotiating and trading - the future of agentization becomes a reality from science fiction. OpenClaw is just the starting point of an acceleration journey. Trillions of dollars are pouring into the AI field. The AI spending of large-scale cloud providers in the United States alone is expected to exceed $650 billion by 2026, which is about ten times the inflation adjusted cost of the Apollo program. Starting from simple chatbots, AI systems are rapidly evolving into fully autonomous agents. These AI agents not only generate content, but also become economic agents - capable of reasoning, acting, trading, debating, coordinating, and without real-time human supervision. The impact of this transformation will be ubiquitous, but the business sector may be the most profound. It is estimated that by 2030, the global consumer business scale of AI agent intervention will reach 3 to 5 trillion US dollars. Even if only 10% of it evolves into programmatic trading between agents, the annual machine native settlement traffic will reach hundreds of billions of dollars. This naturally raises a question: what kind of financial and coordination infrastructure is truly suitable for AI Agent native business? Today's business is designed for humanity, involving identity verification, bank intermediation, legal contracts, settlement cycles, and manual supervision. Our proprietary software cannot open an account at a bank counter, sign documents in person, or wait for ACH to arrive for several days. The infrastructure required for agents must be programmable, always online, globally accessible, permissionless, and machine verifiable by default. Blockchain can meet these constraints, and we have seen this dynamic emerging. During the same period when OpenClaw became popular in January, Solana's trading volume and active addresses also began to climb. The evidence on its AI Agent social network Moltbook suggests that Agent activity may be one of the contributing factors to this growth. X402 is an Internet native payment protocol developed by Coinbase, which allows AI Agent to pay for digital resources in real time without account or complicated high friction authentication process. Since its launch in 2025, its trading volume has continued to accelerate. We are still in the early stages, and today's cases are more directional than decisive. But if investors are excited about the possibility of AI innovation, they should not ignore why we believe that the blockchain track will be the foundation for unlocking a fully autonomous agent world. Many people will correctly point out that today's AI agents do not require blockchain. In recent times, this has indeed been the case, but we believe it is a short-sighted viewpoint. McKinsey recently released a framework that categorizes the degree of AI driven business automation into six levels: from basic subscription assistants (level 0) to fully autonomous agent to agent business (level 5). The core insight is that levels 0 to 4 do not require a new financial track. In every situation, there is a human identity behind the transaction. The user has completed authentication and linked their credit card on ChatGPT, Amazon, or Perplexity. When an agent trades, it is acting as an agent for this human, inheriting their identity, payment credentials, and legal status. The infrastructure for such businesses - shared payment tokens, chargeback systems, fraud detection infrastructure - already exists through Visa or Stripe and operates quite well. The blockchain track becomes critical only at level 5 and above: when an agent trades directly with other agents without human instructions; When there is no human identity to inherit; When payment must be programmatic, conditional, and settled in milliseconds; And when the agent needs a cross platform portable reputation. As long as humans continue to bear economic responsibility, the traditional track is sufficient. Once the agent becomes an independent economic entity, the constraints will be completely changed. To understand where value accumulates and why blockchain is important, we must imagine the logical end state of Agent AI. We are moving towards a world where agents are not just assistants to humans, but independent economic entities. Some are created by companies or individuals, while others are generated by agents themselves, forming increasingly independent systems that can reason, allocate capital, and trade without real-time human supervision. If there were no designated transaction channels by humans (such as going to the bank, using Stripe, or opening a blockchain wallet), the agent would rationally choose the fastest, most reliable, globally accessible, and least frictional and dependent tracks. When the alternative is to open a bank account and wait for ACH settlement during limited banking hours, the agent will naturally choose a blockchain track that operates 24/7 without permission. We believe that there are three core constraints that will drive agents towards the blockchain track: identity and access: how to track the unique identity of AI agents involved in mutual transactions and registration services? What should the new reputation system look like when traditional credit scoring and fraud detection systems are established for humans with physical footprints operating within the jurisdiction? Currency and Payment: What form of currency is needed when an agent is conducting countless micro payments, executing conditional payments, and significantly increasing cross jurisdictional commercial demand? What form of account is required when the agent is unable to enter the bank counter to open an account? Minimizing trust transactions: How AI agents can avoid disputes and frictions that require manual arbitration or other forms of centralized trust - systems they may not be able or willing to access? Identity and Access Before Agent payment, the counterparty must know who - or what - it is dealing with. Traditional identity systems are designed for humans. They rely on government identification documents, physical signatures, and other credentials, assuming that the other end is a legal person. Autonomous AI agents do not have these. It cannot enter the bank to open an account, nor can it sign a contract in a legal sense. However, if we want agents to be able to trade autonomously, they need some way to prove that they are legal and authorized to act. If you connect the Agent to your bank account, the problem will multiply. How do you conduct anti money laundering reviews on software? If the agent acts autonomously, who will bear the responsibility? What if it is manipulated? In simple scenarios, the agent can inherit the owner's credentials (such as ChatGPT Checkout). But this model is not feasible in terms of scale. Multiple agents require separable permissions and consumption limits. It must be possible to isolate problematic behavior without freezing all agents. These scenarios require agents to have their own verifiable identities, rather than borrowing human identities. This is exactly where blockchain based identity comes into play. Through cryptographic techniques, agents can prove that they are authorized to act on behalf of specific individuals or companies without disclosing sensitive information of the parties involved. It can be imagined as a digital authorization letter that can be immediately verified by anyone, anywhere, without the need to contact a lawyer or search a database. Emerging standards such as ERC-8004 of Ethereum propose on chain registry, where agents can establish verifiable credentials and accumulate transaction history and reputation over time. An agent that has successfully completed thousands of undisputed transactions will make a meaningful difference from a brand new agent with no history - and this reputation is cross platform portable. This is important because trust is a prerequisite for business. Merchants have spent years building systems to block robots and crawlers. In the Agent driven economy, they now need to figure out how to release the correct robots. Cryptographically secure and verifiable identity allows merchants to establish trust without the need for human guarantees. Programmable currencies and traditional payment tracks for micro payments are designed for human scale transactions. When you pay for a cup of coffee or a pair of jeans, the credit card fee (usually 2-3% plus about 30 cents per transaction) is negligible. But the commercial operation of agents is on a completely different scale. An agent writing code may initiate 10000 API calls in a single task. An agent comparing prices may need to query hundreds of data providers. The payment needs to be completed within milliseconds, repeated, and the amount is calculated in cents. The credit card network has not been optimized for this behavior. The minimum transaction fee makes micro payments economically unfeasible. The fraud system will freeze accounts exhibiting high-frequency machine behavior. Compared to high-performance blockchain protocols, there is a significant difference in transaction speed. Stablecoins and programmable currencies are truly useful here. On chain transactions can be subdivided into extremely small units, with settlement costs approaching a fraction of cents. More importantly, because payments are programmable, they can be conditional: X is only paid when the API returns valid data; funds are only released when the computing task is completed; As services are consumed, real-time streaming payments are made instead of prepaying an amount that you may not be able to use up. Programmability also enhances capital efficiency. Today, you usually need to pre fund the Agent to access new services, estimate usage, and lock in capital in advance. With smart contracts and on chain collateral, agents can prove their solvency before service delivery without the need to transfer payments. Blockchain has achieved financial infrastructure that matches the working mode of agents: autonomous, high-frequency, conditional, and capital efficient. Traditional businesses embed trust in intermediaries to minimize trust in transactions. Payment processor manages chargebacks. Banks provide settlement guarantees. The court rules on disputes. The contract ultimately relies on the human legal system for execution. When billions of low value transactions occur across multiple jurisdictions, this framework becomes inefficient. An AI agent that trades with another AI agent may not be able to access the legal system of a specific jurisdiction, or may not choose to rely on it. Cross border law enforcement may be slow, expensive, and have uncertain results. Blockchain reduces reliance on these unreliable trust systems by directly encoding and executing logic using smart contracts. For example, smart contracts allow funds to be held in a programmatic manner and released only when predetermined conditions are met. Settlement is deterministic and not affected by the risk of chargeback. The rules are transparent and verifiable in advance for both parties. No need to rely on legal remedies. For large-scale autonomous agents, reducing reliance on centralized intermediaries and manual arbitration can lower friction, improve predictability, and enable programmatic expansion of business. This lower friction infrastructure may expand the uneconomical economic activity boundaries under traditional law enforcement models. Agent businesses supported by blockchain tracks have the potential to accelerate global GDP growth. The question is not whether the Agent business will come, but on what infrastructure it will run on. As AI agents become autonomous economic entities, the number of economic entities in the global economy will grow exponentially. Agents will require a digital native financial track - a technology stack capable of handling programmatic settlements, high-frequency micro payments, permissionless coordination, and minimizing trust identity systems. These principles are the fundamental starting point of blockchain design. We believe that it is reasonable to say that the rapid popularity of AI agents is a strong long-term tailwind for blockchain activities. There is already preliminary evidence on this point, and we believe that most investors underestimate the value creation opportunities contained within it.