Source: Tiger Research; Author: Ekko, Ryan Yoon; Compiler: BitpushNews

An era driven by AI and led by automation is approaching. In order for automation to be truly “autonomous,” it must have native payment capabilities. The market has already begun to prepare for this change.

Core ideas

• Payment subjects are shifting from humans to AI agents, making payment infrastructure a core requirement for true autonomy.

• Tech giants (including Google AP2 and OpenAI delegated payment) are designing approval-based automated payment systems based on existing platform infrastructure.

• Cryptocurrencies (via ERC-8004 and x402) leverage NFT-based identification and smart contracts to enable intermediary-free payment models.

• Tech giants prioritize convenience and consumer protection, while cryptocurrencies emphasize user sovereignty and broader agent-level execution capabilities.

• The key question for the future is whether payments will be controlled by the platform or executed by an open protocol.

1. Payment is no longer exclusive to humans

Source: macstories (provided by Federico Viticci)

OpenClaw has attracted a lot of attention recently. Unlike AI systems such as ChatGPT or Gemini, which are primarily responsible for retrieving and organizing information, OpenClaw enables AI Agents to perform tasks directly on the user's local PC or server.

Through instant messaging platforms such as WhatsApp, Telegram, and Slack, users can issue commands, and Agents autonomously perform tasks including email management, calendar coordination, and web browsing.

Because it runs as open source software and is not tied to a specific platform, OpenClaw functions more like a personal AI assistant. This architecture is favored for its flexibility and user-level control.

However, a key limitation remains: in order for AI Agents to be fully autonomous, they must be able to perform payments. Currently, agents can search for products, compare options and add items to shopping carts, but final payment authorization still requires human approval.

Historically, payment systems have been designed around human actors. In an AI Agent-driven environment, this assumption no longer holds true. If automation is to become fully autonomous, agents must be able to independently evaluate, authorize, and complete transactions within defined constraints.

In anticipation of this shift, major tech giants and crypto-native projects have spent the past year launching technology frameworks designed to enable agent-level payments.

2. Technology giant: Agent payment built on existing infrastructure

In January 2025, Google launched AP2 (Agent Payment Protocol 2.0), expanding its AI Agent payment infrastructure. While OpenAI and Amazon have also outlined plans, Google is currently the only major company with a structured implementation framework.

AP2 divides the transaction process into three authorization layers (Mandate Layers). This structure allows for independent monitoring and auditing of each stage:

1. Intent Mandate: Record what the user wants to do.

2. Cart Mandate: Defines how purchases are performed based on preset rules.

3. Payment Mandate: Performs the actual funds transfer.

Scenario example: How Google AP2 works

Suppose Ekko asks an AI agent on Google Shopping to "find and buy a winter jacket under $200."

• Intent to authorize: Ekko instructs the AI Agent to purchase "a winter jacket with a maximum budget of $200." This information is recorded on-chain as a digital contract and is called intent authorization.

• Shopping Cart Authorization: The AI Agent follows the stated intent, searches partner merchants for products that match "a winter jacket" and "a maximum budget of $200" and adds the qualifying items to the shopping cart.

  "Selected Item: Winter Jacket", "Price Verification: $199 (Budget Fit ✓)"

  Added to cart", "Shipping address confirmed".

• Payment authorization: Ekko confirms the item selected by the AI Agent and clicks the payment approval button. $199 processed through Google Pay. Alternatively, AI Agents can automate payments within predefined parameters.

During the entire process, the user does not need to enter additional information. In the case of Google AP2, the system runs on top of Google Pay and utilizes pre-registered card details and shipping addresses. Because AP2 relies on existing user credentials, it reduces onboarding friction and simplifies the adoption process.

Source: Google

However, Google currently only supports agent-based payments for companies within its partner network. As a result, its use remains limited to a controlled ecosystem, limiting wider interoperability and open access.

3. Cryptocurrency: Self-custody and open exchange

The encryption field is also developing payment infrastructure for AI Agents, but the approach is different from that of major players. While large platforms build trust within a controlled ecosystem, the crypto space starts with a different question: Can AI agents be trusted without relying on a centralized platform?

Two core standards aim to achieve this: Ethereum’s ERC-8004 and Coinbase’s x402.

The combination of identity and payment

Consider the identity layer first. Just like humans need IDs to access digital services, AI Agents running on blockchain networks must be identifiable. ERC-8004 fulfills this function.

It is released as an NFT, but not as a media-like collectible, but as a credential NFT that contains structured identity data. Each token contains three components:

1. Identity

2. Reputation

3. Validation

Together these elements form a verifiable on-chain identity certificate. In e-commerce, participants review ratings and transaction history before transacting, and the same logic applies to AI Agents. ERC-8004 provides agents with verifiable credentials, allowing other agents to evaluate whether transactions are appropriate based on transparent data.

However, identity alone cannot achieve value transfer, a payment mechanism is also required. This role is assumed by x402.

If ERC-8004 is the digital ID, then x402 is the payment rail. Developed by Coinbase, x402 is a crypto-native payment standard for AI Agent. It enables Agents to conduct autonomous transactions using stablecoins.

The core function is automated smart contract execution. Conditional logic such as “automatically transfer money when predefined criteria are met” is embedded directly in the code. Once the conditions are met, settlement occurs without human intervention.

When ERC-8004 for identity and x402 for payments are combined, AI Agents can verify counterparties and execute transactions without relying on a centralized platform. Trust and settlement are handled at the protocol level rather than controlled through the platform.

Scenario example: Inter-Agent commerce based on ERC-8004 and x402

Assume a near-future AI Agent environment: Ekko instructs his AI Agent (Agent A) to purchase a used laptop with a maximum budget of $800. The market runs its own AI Agent (Agent B), which communicates directly with Ekko’s Agent to execute trades.

1. Mutual verification:

   Prior to a transaction, both agents verify each other's credentials and confirm that the product meets specific requirements.

• Identity check: Verified by ERC-8004 NFT

• Ekko's Agent: reputation score 72, confirmed balance $800

• Seller’s Agent: reputation score 70, confirms eligible notebook inventory

• Result: Both Agents are allowed to trade.

2. Smart contract hosting:

   After verification is completed, the transaction begins. Each Agent interacts via the x402 protocol to transfer and confirm funds.

• Escrow: $800 transferred from Ekko’s Agent wallet to the smart contract.

• Conditional Lock: Funds remain locked until receipt is confirmed.

• Release: Upon confirmation of delivery, $800 is automatically transferred to the seller.

3. Settlement and Reputation Update (x402 Settlement and Reputation NFT Update):

   After settlement, the reputation records of both Agents will be updated.

• Ekko's Agent: Reputation 72 → 80 (+5 for fast delivery, +3 for matching description)

• Seller’s Agent: Reputation 70 → 78 (+5 for fast delivery, +3 for matching description)

• The updated evaluation record is written to each Agent's ERC-8004 NFT.

During the entire process, there is no intermediary involved and platform approval is not required. Two AI Agents conduct transactions directly through blockchain-based verification and settlement. This reflects the encryption-native model of Agent-to-Agent commerce.

4. Tech giants vs cryptocurrencies: Differences in areas where AI Agent operates

Control vs Open

Google AP2 represents a controlled model designed for approved partners.

Google limits marketplace participation to vetted merchants, citing consumer protection reasons. Even with a structured authorization framework, Agent behavior cannot be fully guaranteed. Unlike deterministic systems where inputs and outputs are directly matched, the execution of AI Agents produces probabilistic results.

If an Agent connects to an unreliable partner and a transaction error occurs, the responsibility may ultimately fall on the payment infrastructure provider. To reduce the probability of failure by even 0.01%, Google has an incentive to shrink its ecosystem. This restricted ecosystem increases stability and regulatory capabilities, but may also limit the Agent's ability to operate autonomously in the broader market and optimize across multiple options.

In contrast, ERC-8004 and x402 reflect a more open architecture. Encryption mode is designed to be permissionless and interoperable, rather than tied to a platform.

Efficiency and use cases

AI Agent is still in the early stages of development. End-to-end execution from complex requests to autonomous payments is not yet seamless. However, the expected long-term scenario is for the Agent to independently manage daily consumption. For example, a user might instruct an Agent to restock groceries, and the Agent assesses inventory gaps and automatically completes the purchase.

Large platforms may try to aggregate major retail channels to support this model in a unified environment. This approach enables reliable everyday usage scenarios within a controlled framework. However, closed ecosystems face structural limitations in integrating all potential counterparties, including small online merchants, independent websites, decentralized finance protocols, and trading venues.

Additionally, if digital content increasingly shifts to a paid-access model, Agents may need to perform high-frequency micropayments. Open encryption standards may have structural advantages. For example, an AI Agent could purchase 1,000 creator-generated images for $0.01 per unit, or pay $1 for access to a research article. For small, programmable payments, crypto-native rails may offer greater operational efficiency.

That said, the lack of a centralized institution also brings trade-offs. Identity assessment criteria must be established in a decentralized manner, with no single entity bearing ultimate responsibility for failure. Striking a balance between openness and accountability remains a key design challenge that will depend on advances in technology maturity and ease of use.

Summary

Both major technology companies and the encryption field are pursuing the same goal: realizing autonomous AI Agent business. The difference is architecture. The big players favor closed, controlled systems, while the crypto space is pushing for open, protocol-based models.

This is not a zero-sum game, and the more likely trajectory is interoperability between the two approaches. In the current stage of technological advancement, continued development must prioritize reliability and user experience.