A Chainalysis report shows AI agents made over 138 million crypto payments, predominantly small transactions, highlighting a growth in agentic commerce.
AI Agents Drive Surge in Crypto Payments
A recent Chainalysis report highlights the growing trend of autonomous AI agents engaging in cryptocurrency transactions, revealing that they have executed more than 138 million payments through Coinbase's x402 protocol. This form of commerce, termed "agentic commerce," is rapidly emerging as a significant use case within the crypto ecosystem.
The report indicates that these transactions are predominantly small, with over 87% falling between one-hundredth of a cent and 10 cents. Such micro-transactions are ideal for stablecoins, which can handle these amounts more efficiently than traditional credit card systems.
Significant Growth in AI Payment Activity
The x402 protocol, part of Coinbase's layer-2 network, has seen a remarkable increase in payments, surging from approximately 1,500 in mid-2025 to over 138 million by the third quarter of 2026. This growth is attributed to developers allocating budgets for AI agents to purchase computing resources, real-time web data, and coding tools on demand.
The report reveals that U.S.-based wallets are the primary drivers of this activity, accounting for 88.1% of identifiable payments related to AI model inference. In contrast, Chinese wallets represented only 0.1%, although they were significant in spending on security and code-analysis services.
Maturing Infrastructure for Agent Payments
The findings come as the infrastructure supporting AI agent payments evolves rapidly, with Visa, Mastercard, and Stripe now facilitating these transactions. Coinbase has introduced tools that enable agents to trade and transact, while Google has launched a dedicated protocol for agent payments.
BlackRock has identified autonomous agents as a potential new source of demand in the crypto market. Will Su, the firm's head of digital assets research, noted the potential for AI agents to generate ongoing demand for frequent, low-value payments.
However, caution is advised regarding the interpretation of these figures. Previous research by TRM has suggested that many transactions attributed to agents may actually involve scripts or self-dealing, indicating that the volume of transactions may not accurately reflect genuine agentic commerce activity.
Despite these considerations, the trend suggests a shift towards a future where the next wave of crypto users may not be human at all, but rather autonomous agents operating in the digital economy.
