Agent-to-Agent Commerce - Agent-to-Agent Business

Basic Information

• Domain: AI Business / FinTech
• Type: Emerging Business Model
• Development Stage: Transition from Experimentation to Mainstream (2025-2026)
• Key Players: Visa, Coinbase, Google, IBM, Major Retail and Financial Enterprises

Concept Description

Agent-to-Agent Commerce / Agentic Commerce refers to the full-process business activities where AI agents autonomously complete the discovery, comparison, negotiation, purchase, and payment of goods and services on behalf of users. In the most advanced scenarios, the AI agents of sellers and buyers can directly negotiate transactions without human involvement at every step.

Milestone Events

Visa Agent Transaction Experiment (December 2025)

• Successfully completed hundreds of agent-initiated transactions in a real-world environment
• Transactions were initiated, authenticated, and completed entirely by AI agents—no human checkout button required
• Visa Statement: "2025 will be the last year consumers shop and checkout alone"

Trusted Agent Protocol (October 2025)

• An open framework launched by Visa in collaboration with 10+ partners
• Designed based on existing web infrastructure
• Helps merchants distinguish between malicious bots and legitimate AI agents representing consumers
• Enables secure agent-driven checkout

Coinbase Agent Wallet (February 2026)

• Wallet infrastructure specifically built for AI agents
• Supports autonomous spending, earning, and trading
• Bridges the connection between cryptocurrencies and agent commerce

Market Data

• IBM 2026 Survey: 45% of consumers have used AI in some part of their purchasing journey
• Limiting Factor: Not AI capability, but whether merchants can verify agent identity and authorization scope
• Chinese Tech Giants: Actively shifting to compete in the agentic commerce space

Technical Infrastructure

• A2A Protocol (Google): Standard for agent-to-agent communication
• Trusted Agent Protocol (Visa): Identity verification for agent transactions
• Agentic Wallets: Payment infrastructure dedicated to agents
• MCP Protocol: Integration standard for agents and merchant systems

Application Scenarios

• Automatic Shopping: AI agents complete daily shopping on behalf of users
• Service Booking: Automatic booking of hotels, restaurants, and repair services
• Price Negotiation: Automatic price comparison and negotiation between agents
• Supply Chain Procurement: Enterprise agents autonomously complete procurement processes
• Financial Transactions: Automatic investment and trading by AI agents

Regulatory Challenges

• Existing consumer protection laws designed for humans may slow the development of agent commerce
• Lack of legal frameworks for agent identity verification and authorization confirmation
• Issues of liability in agent transactions
• Preventing agents from being used for fraud or market manipulation
• Compliance of cross-border agent transactions

Key Conclusions from the 2026 Shoptalk Conference

• AI, agentic commerce, and the future of retail were core topics
• Agent commerce will scale first in areas where trust and authorization are clearly established
• The retail industry faces fundamental changes driven by AI agents

Relationship with the OpenClaw Ecosystem

Agent-to-Agent commerce provides OpenClaw with significant commercialization scenarios. OpenClaw's personal AI agents can execute shopping, booking, and payment tasks on behalf of users. The platform needs to integrate identity verification frameworks like Visa's Trusted Agent Protocol to ensure the security and legality of agent transactions. This is also a key capability for OpenClaw to upgrade from "information assistant" to "action agent."