Author: RWA Research Institute

In March 2026, Peter Schroeder, Head of Global Markets at Circle, released a set of data on platform X: In the past nine months, AI agents completed 140 million payments with a total transaction value of 43 million USD. Among these, 98.6% were settled in USDC, with an average transaction amount of only 0.31 USD. More importantly, the number of AI agents with purchasing power has exceeded 400,000.

This set of data speaks louder than any financial report: AI agents are transitioning from concept to real economic activity.

400,000 AI agents, 140 million transactions, 43 million USD—this is a value exchange autonomously completed between machines. No human intervention, no bank approvals, no credit card validations. Between code and code, between protocols and protocols, the processes that previously required human signatures, reconciliations, and settlements have been completed.

Circle's stock price has risen from 60 USD to 105 USD over the past few trading days, an increase of 75%. The market interprets this rise as a positive response to the financial report—Circle achieved a revenue of 770 million USD in Q4 2025, a year-on-year growth of 77%, and a net profit of 133 million USD. However, what truly deserves attention is not the numbers themselves but the structural changes behind them: as AI agents become new economic entities, the entire logic of financial infrastructure needs to be rewritten.

In this rewriting process, a deeper question is emerging: When AI agents begin to have disposable funds, when they can earn USDC by completing tasks, how will they manage these funds? Payment is the first step; asset management is the second step. What RWA (Real World Assets) needs to address is this second step.

1. From Payment Capability to Asset Holding

To understand what financial services AI agents need, we must first understand their economic activity patterns.

Deloitte pointed out in its "2026 Technology, Media, and Telecommunications Industry Predictions" report that if companies and service providers can achieve efficient collaborative scheduling of intelligent agents, the global agent-based AI market is expected to reach 45 billion USD by 2030. The basic characteristic of this multi-agent collaboration model is: a complex task is divided into multiple steps, completed by different specialized agents, with each call accompanied by a micro-payment.

Take API calls as an example. An AI application may need to simultaneously call multiple large language models, access multiple databases, and use several computing resources. Each call accumulates amounts of 0.01 USD, 0.05 USD, or 0.1 USD. These payment amounts are very small, but the frequency is very high. Circle's data shows that in the past nine months, there have been 140 million transactions, averaging just 0.31 USD per transaction—this is a typical feature of the micropayment market.

But the problem is, when AI agents continuously generate income—whether by providing services to users or by participating in distributed computing networks—they will accumulate funds in their accounts. These funds cannot remain in a liquid state indefinitely. Any rational economic entity will consider: how to handle idle funds?

This is the logical starting point for AI agents to transition from being "payers" to "asset holders."

In the traditional financial system, individuals and enterprises would deposit short-term idle funds into banks, purchase money market funds, or short-term government bonds to earn returns. AI agents also need such capabilities—not for speculation, but to optimize their own economic models. It is necessary to keep a portion of USDC in the account for payments, but if the amount exceeds a threshold and just sits idle, it means a loss of opportunity cost. If surplus funds can be automatically invested in a tokenized fund backed by short-term U.S. Treasury bills and automatically redeemed when needed, then its "operational efficiency" is improved.

Further, if AI agents need to reserve value for long-term operations or hedge against cost uncertainties caused by gas fee fluctuations, they may develop a demand for allocating assets of different risk levels. At this point, they are no longer just "payers," but "investors"—even if this investor is a piece of code.

Circle addresses the issue of making AI agents "payers." To make them "investors" requires another set of infrastructure.

2. RWA and AI Agents: A Mutual Pursuit that is Happening

What Circle has done over the past few years can be summarized as building three layers of capabilities.

The first layer is the issuance of stablecoins and liquidity networks. According to Circle’s official disclosures, by the end of 2025, the circulating supply of USDC is expected to reach 75.3 billion USD, a year-on-year growth of 72%, with a share of nearly 50% in the stablecoin trading volume. This provides a usable value carrier for AI payments.

The second layer is an efficient on-chain settlement network. In August 2025, Circle launched the Arc chain specifically for institutional financial services. In March 2026, Circle introduced the Nanopayments system, which aggregates thousands of micro-payments off-chain and periodically packages them on-chain, lowering transaction costs for developers to zero. The test network already supports 12 EVM chains, including Arbitrum, Arc, Avalanche, Base, and Ethereum. At the payment protocol level, the x402 protocol allows websites or APIs to issue HTTP 402 payment requests directly when returning requests, embedding payments directly into internet requests.

The third layer is the connection to the traditional financial system. The Circle Payments Network (CPN) connects banks, payment service providers, cross-border clearing institutions, and corporate clients. As of February 2026, 55 financial institutions had joined, with an annual transaction volume of about 5.7 billion USD. In February of this year, new direct payment systems for local currencies and stablecoins were added in multiple regions, including Asia and the Middle East.

These three layers of capability make up the "payment infrastructure" for the AI agent economy. However, a complete economic system also requires "asset management infrastructure," where RWA can enter.

The exploration of tokenization of RWA (Real World Assets) in recent years has mainly focused on "on-chain mapping" in traditional finance. According to Defillama data, by June 2025, the total locked value (TVL) of RWA reached 12.5 billion USD, a growth of 124% compared to 2024. Global leading banks like Citibank and Standard Chartered are exploring use cases for RWA in payment settlement, asset management, and cross-border transactions.

However, to enter the economic world of AI agents, RWA needs to undergo an "AI native" transformation. This is not just a matter of putting assets on-chain; it is about making assets "understandable by AI and tradable by AI."

First is data standardization. Leading RWA projects like Ondo Finance are pushing to turn underlying cash flows, legal terms, and risk ratings into structured, machine-readable data formats. In July 2025, Ondo Finance launched a tokenized U.S. Treasury bond project for global investors, which was included in the White House report published by the U.S. President's Digital Asset Market Working Group.

Next is logic programmability. Rules for dividends, interest payments, buybacks, settlements, etc., are written into smart contracts and executed automatically by code. The interaction between AI agents and assets can achieve "trustlessness"—there is no need to trust that the counterpart will perform, only to trust that the code will operate according to established rules.

The third is liquidity fragmentation. After RWA is tokenized, it can theoretically be split into extremely small units—0.01 USD of government bonds, 0.1 square meters of real estate income rights—this is crucial for the small allocation needs of AI agents. Nanopayments have already proven that micropayments are technically feasible, and the same logic can extend to micro-investments.

JP Morgan's Kinexys division provides a reference case. In May 2025, Kinexys completed the first public transaction of tokenized U.S. Treasury bonds on the Ondo Chain test network, using Ondo Finance's tokenized U.S. Treasury bond fund (OUSG), and settled using Chainlink's cross-chain infrastructure. The transaction followed the "Delivery versus Payment" (DvP) model, achieving a simultaneous exchange of assets and payments. JP Morgan's Kinexys division currently processes over 2 billion USD in transactions daily and has facilitated over 15 trillion USD in nominal value transactions since its establishment.

The value of this case lies in: it demonstrates the combination of RWA and institutional-level payment settlement networks. In the future AI agent economy, the trading entity may shift from JP Morgan to an AI agent, and the transaction scale may change from millions of dollars to just a few dollars, but the underlying logic remains the same—value transfer and value storage need a seamless connection.

3. Beyond the Payment Network, There is Another Layer of Imagination

If we connect the above logic, a complete closed loop begins to emerge:

An AI content generation agent accumulates a significant USDC balance by providing services to multiple clients. Its underlying protocol sets cash management rules: the part of the balance exceeding 1,000 USDC is automatically allocated through an RWA aggregator, averaged across three tokenized short-term treasury funds and one tokenized green energy fund. When customer demand decreases in a certain month and the account balance needs to be replenished, the protocol automatically redeems part of the RWA shares for USDC for daily operations.

In this process, the actions completed by the AI agent include: monitoring account balances, assessing the risk-return characteristics of different assets, executing purchases and redemptions, and recording transaction flows for subsequent audits. All actions are completed automatically by code without human intervention.

For example, an AI travel planner books flights and hotels for users, and the user transfers a sum of USDC to its account as a budget. While waiting for the flight, the AI agent detects an RWA insurance product based on flight delay data is being offered. It uses a portion of USDC that is temporarily idle in the account to automatically purchase a micro-share of that insurance. Hours later, the flight is delayed, and the RWA insurance product automatically triggers compensation according to the rules, increasing the AI agent's account balance.

Each technical module that constitutes these scenarios already exists: USDC provides the value carrier, Nanopayments resolves the micropayment cost issue, the x402 protocol enables payments to be directly embedded in internet requests, tokenized treasury bonds are running on platforms like Ondo Chain, and the DvP settlement mechanism has been validated by JP Morgan. The remaining work is to integrate—link the payment layer, asset layer, and transaction layer together, allowing AI agents to call these financial functions like invoking an API.

Li Ming, Executive President of the Hong Kong Web3.0 Standardization Association, noted when commenting on RWA development, "We hope to find a standardized entry point for Web3.0 that can connect the RWA ecosystem." For the AI agent economy, this entry point may be precisely the connection point between payment and assets.

4. The Old Problem of the New World: Risk and Responsibility

Of course, there are several obstacles to overcome from today’s AI payments to tomorrow’s AI asset management.

First is the issue of data authenticity. The underlying assets of RWA are off-chain, and their status, value, and risk information need to be reliably transmitted on-chain. If AI agents rely on incorrect or tampered data, their "investment decisions" will be flawed. The "RWA Industry Development Research Report" released by organizations including the Hong Kong Web3.0 Standardization Association points out that assets that achieve scaled deployment must meet three major thresholds: value stability, clear legal rights, and verifiable off-chain data.

Secondly, there is the model risk of AI agents. Even if the data is accurate, the investment decision logic of AI agents may still fail. Who is responsible for the erroneous decisions made by AI agents? Is it a person, the protocol, or the AI agent itself? This issue of responsibility attribution currently lacks answers in both legal and regulatory contexts.

The third is liquidity risk. The on-chain trading depth of RWA is far from that of mainstream cryptocurrencies, and some assets may have poor liquidity. When a large number of AI agents need to redeem the same RWA fund simultaneously, the certainty of successful transactions is uncertain.

The fourth is regulatory differences. Different countries have varying regulatory attitudes towards RWA, and the legal status of the same asset can differ drastically across jurisdictions. AI agents need to be able to recognize and handle this complexity, which poses high requirements for current AI capabilities.

Finally, there is the risk of technical security. Risks such as smart contract vulnerabilities, cross-chain bridge attacks, and private key leaks do not disappear just because the trading entity is AI. On the contrary, when AI agents achieve automated trading, the speed and scale of exploiting vulnerabilities may far exceed that of manual operations.

Conclusion

Returning to the opening data: 400,000 AI agents, 140 million transactions, 43 million USD.

The significance of these numbers lies not in the scale itself—compared to the trillions of dollars in payments made by humans each year, 43 million USD is negligible. Their real significance lies in revealing a direction: machines are becoming independent economic entities, possessing their own income, accounts, and payment capabilities.

When machines have income, they will soon have asset management needs. This is not a distant imagination, but the natural path of evolution for the AI agent economy.

Circle is laying the groundwork for this future by creating a "payment nervous system"—enabling AI agents to transfer value efficiently and at low cost. What the RWA sector needs to do is to become the "energy storage system" of this economic entity—allowing AI agents to manage their assets as they manage their code.

If this judgment holds true, then the question RWA practitioners need to consider today is: When 400,000 AI agents begin to seek configurable assets, when asset management needs arise after 140 million payments, are you ready for your RWA products to be evaluated, chosen, held, and traded by AI agents?

Related Reading: Circle's Turning Point: Stock Price Doubling, On-chain Trading Crushing USDT, Precisely Positioning Agent Payments

(This article is based on Circle's official financial reports and announcements, Deloitte's "2026 Technology, Media, and Telecommunications Industry Predictions," Defillama data, Ondo Finance public information, JP Morgan Kinexys official disclosures, and the Hong Kong Web3.0 Standardization Association's "RWA Industry Development Research Report," and does not constitute any investment advice. The market has risks; investment should be cautious.)

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。