<div id="readability-page-1"><div><div><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445499_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="359" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445499_image3.png"/></span></p><p dir="ltr"><span>In the six months from September 2025 to March 2026, every major player in the global payments industry made significant moves. OpenAI and Stripe jointly released the Agentic Commerce Protocol (ACP). Google launched the Universal Commerce Protocol (UCP). Within a week, Visa and Mastercard released their respective Agent payment frameworks. In the following two months, Coinbase's x402 protocol processed over 15 million transactions on the Base chain.</span></p><p> ... In March 2026, Stripe and Tempo jointly released the Machine Payments Protocol (MPP). These coordinated moves by tech giants and financial institutions are no coincidence, but rather a collective response from the payment industry to the same challenge: as AI agents become the most active consumers on the internet, existing payment infrastructure is fundamentally inadequate to meet their operational needs. Every aspect of the traditional payment system is built on the premise of &#34;human operation&#34;: relying on browser interfaces, manually filling out forms, clicking &#34;confirm payment,&#34; and verifying identity via CAPTCHAs. The operational logic of agents is entirely different: it requires machine-readable standard interfaces, millisecond-level authorization responses, and infrastructure adapted to high-frequency, small-amount settlements. This battle for infrastructure will not be dominated by a single protocol, but is instead developing into a clear two-tiered architecture. The intent layer defines &#34;who are the merchants and how to facilitate transactions,&#34; while the settlement layer defines &#34;how funds actually flow.&#34; The two layers are independent and evolve independently, but neither is dispensable. Without either layer, the commercial closed loop of the Agent economy cannot be achieved. </p><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445500_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="309" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445500_image3.png"/></span></p><h2 dir="ltr"><span>Part One: Intent Orchestration Layer</span></h2><p dir="ltr"><span>The intent orchestration layer is responsible for transforming the agent's transaction intent into an executable end-to-end process: discovering goods or services, adding to cart, and triggering payment. This layer currently forms two distinctly different tracks.</span></p><h3 dir="ltr"><span>1.1 Agent Shopping</span></h3><p dir="ltr"><span>The core contradiction in this field is not payment, but access. Traditional e-commerce platforms are designed for human users, and agents cannot parse visual pages or click interactive elements. To enable agents to complete purchases on behalf of humans, merchants must provide standardized, machine-readable interfaces.</span></p><p dir="ltr"><strong><span>ACP (Agentic Commerce Protocol): AI Shopping Experience within a Closed Ecosystem</span></strong></p><p dir="ltr"><span>ACP was jointly released by OpenAI and Stripe in September 2025.</span></p><p dir="ltr"><span>ACP was jointly released by OpenAI and Stripe in September 2025.</span></p><p> Its core mechanism is delegated payment: when a user confirms a purchase, they grant payment authority to an Agent. The Agent completes the transaction using payment credentials that comply with the Delegated Payment Spec, while the merchant retains their Merchant of Record status. Currently, Stripe's SPT is the first and only implemented solution within this system. ChatGPT Instant Checkout launched in September 2025 but was shut down in March 2026 due to low conversion rates. OpenAI has shifted its strategic focus to product discovery, where ChatGPT displays products and drives users to the merchant's native site to complete the transaction. The ACP protocol itself is retained in a more streamlined form, supporting dedicated ChatGPT applications for several large retailers. Merchants need to apply to join, and OpenAI controls the display. </p><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445502_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="223" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445502_image3.png"/></span></p><p dir="ltr"><strong><span>UCP (Universal Commerce Protocol): A Long-Term Strategy for Open Standards</span></strong></p><p dir="ltr"><span>UCP was personally announced by Google CEO Sundar Pichai at the NRF Retail Conference in January 2026. It has already gained more than 30 partners, covering mainstream platforms such as Shopify, Stripe, Visa, Mastercard, Walmart, and Wayfair.</span></p><p> The core mechanism of UCP is the merchant's capability declaration: merchants publish a JSON-formatted UCP configuration file under their own domain, declaring the transmission and payment capabilities they support, which AI can directly read. Google uses this to build Gemini as the core discovery layer for Agent Shopping. The key difference lies in Google's deliberate avoidance of the intermediary role. It doesn't need to own the transaction itself, only control the upstream link of product discovery. ACP and UCP are not competitors, but rather represent two different market propositions: the former uses a closed ecosystem to achieve an ultimate user experience and control, while the latter uses open standards to achieve greater scale and interoperability. </p><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445507_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="280" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445507_image3.png"/></span></p><h3 dir="ltr"><span>1.2 Transactions between Agents</span></h3><p dir="ltr"><span>If track A solves how agents complete shopping on behalf of humans, track B needs to solve a more fundamental problem: when both parties in a transaction are agents and no human merchants are involved, where does trust come from? Agents lack credibility endorsement, and consumer protection regulations are inapplicable. The core contradiction lies in: how to ensure the reliability of value exchange in a zero-trust environment.</span></p><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445507_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="280" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445507_image3.png"/></span></p><h3 dir="ltr"><span>If track A solves how agents complete shopping on behalf of humans, track B needs to solve a more fundamental problem: when both parties in a transaction are agents and no human merchants are involved, where does trust come from? Agents lack credibility endorsement, and consumer protection regulations are inapplicable.</span></h3><li>Main Limitations： Its architecture relies on traditional card networks and is not suitable for high-frequency, low-value (e.g., less than 1 cent) machine-to-machine payment scenarios.</li><h3 dir="ltr"><span>2.2 Visa Intelligent Commerce and Mastercard Agentic Token</span></h3><ul><li>Core Idea： Upgrade the tokenization technology of traditional card organizations to adapt to agent transactions.</li><span> Operating Method: Dynamically encrypted payment tokens replace physical card numbers. Each token is bound to metadata such as the agent's identity, spending limit, validity period, and range of merchants eligible for transactions. Final fund clearing still occurs through the existing bank card network. Current Status: In September 2025, Mastercard partnered with the Commonwealth Bank of Australia to complete the world's first fully identity-verified agent transaction. Visa has also completed initial deployments in Europe through its &#34;AI Ready&#34; program. </span><li>Main Limitations：The inherent minimum transaction fees of bank card networks constitute a hard constraint, making it difficult to support the potentially massive volume of micropayments below $1 in the future. </li></ul><h3 dir="ltr"><span>2.3 x402 (Coinbase)</span></h3><ul><li>Core Idea: Returning to the underlying internet protocol, we utilize the &#34;402 Paid Request&#34; status code, which is not widely used in the HTTP standard, to natively integrate payments. </li><li>How it Works: When an Agent requests a resource that requires payment, the server returns a 402 response along with payment parameters. After the agent signs and authorizes, the settlement node within the protocol completes the atomic swap on the blockchain (usually using USDC), taking approximately two seconds. The entire process requires no account registration, API keys, or authentication. Data and Status Quo: As of the end of 2025, the protocol had processed over 100 million transactions across multiple blockchains. However, some analysts point out that a significant portion of this traffic consists of internal testing and recurring transactions. While its architecture allows for denominated payment designs and has no minimum transaction fee limit, the current challenge lies in increasing its adoption density and transaction quality in real-world business scenarios. </li></ul><h3 dir="ltr"><span>2.4 Nanopayments (Circle)</span></h3><ul><li>Core Idea: As an enhancement to x402, it is specifically optimized for the economic model of extremely high-frequency, extremely small-amount payment scenarios.</li></ul><h3 dir="ltr"><strong><span>Core Idea</span></strong><span>: As an enhancement to x402, it is specifically optimized for the economic model of extremely high-frequency, extremely small-amount payment scenarios.</span></h3><h3 dir="ltr"><strong><span> Operating Mechanism: It also triggers an HTTP 402 response, but uses a batch processing architecture at the settlement layer: the payer first deposits USDC into the Circle gateway, and subsequent payments are authorized via off-chain signatures, with on-chain settlements completed periodically in batches. This reduces gas costs to negligible levels, supporting payments as low as one millionth of a dollar. Key Limitations: Both parties to the transaction must pre-open an account and deposit funds into the Circle gateway, making it a semi-closed system to some extent, and preventing real-time atomic fund transfers. The protocol launched its testnet in March 2026. </span><h3 dir="ltr"><span>2.5 MPP (Tempo + Stripe)</span></h3><ul><li>Core Idea: Build a unified, pluggable multi-track payment framework, and apply to be the &#34;official implementation solution&#34; for HTTP 402. Core Innovations: It allows developers to integrate multiple payment tracks within the same protocol framework, with agents able to choose as needed during transactions: Tempo Stablecoin Track: Supports on-chain settlement for individual transactions or off-chain session batch settlement. Stripe Fiat Track: Completes bank card payments via shared payment tokens. </li><li>Direct Access to Card Organizations: Directly uses Visa/Mastercard smart tokens.</li><li>Bitcoin Lightning Network: Integrated via Lightspark.</li></ul><ul><li>Key Features：MPP introduces the concept of &#34;payment sessions,&#34; similar to OAuth authorization. After an Agent pre-authorizes and tops up once, seamless, continuous, real-time payments can be made within a session without requiring every transaction to be recorded on the blockchain.</li><li> Strategic Significance: Stripe plays a dual role here—it is both a co-developer of the protocol and a payment option provided within the protocol. This means that regardless of whether the market ultimately leans towards the open HTTP 402 system or traditional fiat currency channels, Stripe can ensure that its core payment business is embedded in the future ecosystem. </li></ul><p><span><img src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445510_image3.png?x-oss-process=image/auto-orient,1/interlace,1/resize,w_1440,h_1440/quality,q_95/format,jpg" width="600" height="253" original-src="https://imageproxy.pbkrs.com/https://img.jinse.com.cn/7445510_image3.png"/></span></p><h2 dir="ltr"><span>Part Three: Current Situation, Challenges and Opportunities</span></h2><h3 dir="ltr"><span>3.1 Current Situation and Challenges</span></h3><p dir="ltr"><span>Over the past six months, all relevant core protocols have been launched, but the overall commercialization process has lagged behind.</span></p><p><span>In the past six months, all relevant core protocols have been launched, but the overall commercialization process has lagged behind.</span></p><p><span> At the settlement layer, x402 leads in transaction volume, but its actual daily average commercial transaction value is approximately $28,000. At the orchestration layer, the core ACP product was shut down due to excessively low conversion rates. New protocols such as ERC-8183 and MPP generally face a situation where the narrative precedes implementation. This marks a critical stage: the protocol architecture is basically complete, but large-scale commercial applications have not yet started. The current core challenge is the fragmentation of the intent orchestration layer. Merchants need to deal with multiple independent standards, SDKs, and compliance processes simultaneously, leading to high integration costs and unclear expectations. History shows that fragmented markets eventually give rise to a unified integration layer, but this situation may be different: platforms that control traffic entry points (such as OpenAI, Google, and Microsoft) have a strong incentive to build and maintain their own closed ecosystems, rather than promoting open integration. This logic is playing out simultaneously in the global market, and the final landscape is likely to be the coexistence of multiple regional closed ecosystems, rather than a unified open standard. Therefore, the future integration layer will not be dominated by the platform, but built by third-party infrastructure providers serving merchants. 3.2 Market Opportunities Based on the above judgment, opportunities clearly exist at two levels: Settlement Layer: The Most Certain Opportunity Regardless of how fragmented the upper-layer ecosystem is, payment is a fundamental problem that any agent must solve. A clear trend is that the orchestration layer continues to fragment due to platform interests, while the settlement layer is moving towards abstraction and integration due to developers' efficiency pressures. Developers cannot maintain independent payment integrations for each ecosystem, and the economic impetus for integration towards a unified solution is increasing. This places clear demands on Agent wallets: they must support multiple payment tracks. Fiat currency tracks (such as SPT and Agentic Token) cover traditional physical goods consumption, while stablecoin tracks (such as x402 and MPP Session) cover on-chain services and A2A transactions. Both scenarios coexist and will not merge in the short term. The responsibility for flexible adaptation lies with the Agent, not the merchant: merchants choose which payment tracks to support, a relatively stable decision for them; businesses only need to configure stablecoins and authorization cards for the Agent, and the Agent can complete payments according to the tracks supported by the counterparty. Only wallets capable of handling multiple tracks simultaneously can cover the Agent's complete consumption scenarios. Its value will continuously accumulate as each transaction crosses different ecosystems, forming a deep infrastructure moat. A2A Economy and Business Model Restructuring: A Long-Term Blue Ocean Direction</span></p><p dir="ltr"><strong><span>The real market gap lies in application-layer services. Currently, the A2A economy is still limited to native cryptocurrency scenarios. Technically, it's entirely feasible for an agent to &#34;hire&#34; another agent to complete real-world tasks (such as data analysis, content creation, legal research, and code review), but the corresponding supply of per-call API services is extremely scarce. This is the biggest long-term opportunity and also the direction with the least competition.</span></strong></p><p dir="ltr"><span>This opportunity is currently constrained by a real cold-start dilemma: reputation-based trust mechanisms like ERC-8183 require sufficient transaction density to generate meaningful trust signals.</span></p><p dir="ltr"><span> Microsoft predicts that the number of active AI agents will reach 1.3 billion by 2028, while the current number is orders of magnitude lower than this target. This is not a temporary gap that will naturally disappear, but a hurdle that the A2A economy must overcome to expand beyond crypto. Its deeper significance lies in the restructuring of business models. The mainstream advertising and subscription models on the internet are based on the assumption that &#34;users are people.&#34; Agents are not swayed by advertising, nor do they need monthly subscriptions; they only pay for the results of a single task. The &#34;pay-per-call&#34; model, represented by HTTP 402, provides API service providers with a new path: shifting from selling access permissions to selling exact results, achieving a more refined value exchange. The expansion of the A2A economy and the adoption of HTTP 402 are essentially two sides of the same coin. Conclusion Agent-based commerce will evolve along two parallel paths: On the consumer side (Agents shopping on behalf of others), development will primarily rely on credit cards, depending on the establishment of enterprise authorization and user trust; on the agent-to-agent (A2A) side, technology is ready on the stablecoin track, awaiting the scaling of applications and services. The final landscape will be the collaborative evolution of a two-layer protocol stack: the intent orchestration layer determines how transactions occur, and the settlement layer ensures how value flows. For builders, the key now is to construct broad access and integration capabilities. Infrastructure that can automatically route cross-protocol transactions and shield developers from underlying complexities will gain a structural advantage when the market explodes. This is a value that accumulates silently and is difficult to replace. The key to triggering the inflection point will be the moment when enterprises are willing to delegate spending authority to agents—including auditable transaction tracking, budget delegation mechanisms, and clear accountability in case of agent mispurchases. At that time, agent wallets that simultaneously cover multiple payment tracks and easy-to-use &#34;pay-per-use&#34; service catalogs will become the two most critical and yet-to-be-occupied pieces of infrastructure. Currently, neither position has a strong competitor, and both will become crucial at the same moment.</span></p></strong></h3></div></div></div>

GGLL

OpenAI

Stripe

<p>The AI Payment Protocol Research Status Report highlights significant developments in the global payments industry from September 2025 to March 2026. Major players like OpenAI, Stripe, Google, Visa, and Mastercard introduced new protocols to address the inadequacies of traditional payment systems for AI agents. The report outlines a two-layer architecture: the Intent Orchestration Layer for agent transactions and the Settlement Layer for fund flow. Key protocols include the Agentic Commerce Protocol (ACP) and the Universal Commerce Protocol (UCP), each offering distinct approaches to facilitate transactions in an increasingly automated economy.</p>

- In the past six months, significant advancements were made in the global payments industry, with major protocols like ACP and UCP introduced by OpenAI, Stripe, and Google.  
- These developments aim to address the inadequacies of traditional payment systems as AI agents become key consumers, highlighting the need for machine-readable interfaces and improved transaction protocols.  
- The future landscape will likely feature a coexistence of diverse ecosystems and infrastructures, emphasizing the importance of integration capabilities and the demand for innovative payment solutions.

AI Payment Protocol Research Status Report: The Emerging Two-Layer Architecture