Featured image: The image shows a server room with multiple server racks, each with blue LED lighting. AI-generated image; not a photograph of the reported event.
AI agents are now making direct crypto payments for digital services. Over 138 million transactions have been recorded through Coinbase’s x402 protocol on the Base blockchain.
AI Agents Begin Paying for Services
These payments cover access to AI models, data, and computing power. The x402 protocol lets AI agents pay automatically without human input.
Chainalysis found most transactions are very small. This range is too low for traditional card networks to process efficiently.
Over 87% of analyzed payments fall into this low-value range. This highlights stablecoins’ ability to support micropayments.
U.S. Leads, China Focuses on Security
U.S.-linked wallets make up a majority of AI model inference payments. Chinese-linked wallets hold a small share of such payments.
Chinese wallets, however, account for a significant portion of payments for security and code-analysis tools. Much of this activity targets AgentLisa, an AI-powered smart contract auditing platform.
Eric Jardine of Chainalysis said one possible reason for the gap is that Chinese developers often use open-weight models like DeepSeek and Qwen. These models are free and don’t require paid inference services.
“That’s one possible interpretation, but we wouldn’t go so far as to say open-source AI is starving Asia’s on-chain inference market.”
These figures only include payments through x402. They don’t reflect the full global AI payment market or prove real commercial demand exists.
How x402 Enables AI Spending
x402 is a protocol developed by Coinbase. It lets AI agents pay for digital services automatically using crypto.
It supports continuous, real-time transactions—ideal for AI systems that run constantly.
Traditional card systems aren’t built for tiny payments. Stablecoins, however, offer low volatility and fast transactions. They suit AI agents that need reliable, frequent spending.
AI agents may use stablecoins to buy data or run computations. While each transaction is small, millions of them can add up to meaningful value.
Yat Siu of Animoca Brands estimates up to 100 billion AI agents could be online in the coming years. He believes the next billion crypto users may be AI agents, especially in supply chains, content, or infrastructure.
Stephanie Cohen of Cloudflare says stablecoins are ideal for early AI spending. They offer low cost, speed, and stability—key for autonomous systems.
AI Agents as Economic Actors
These transactions show AI agents are becoming active participants in digital economies. Each agent acts like a new blockchain user.
While 138 million transactions occurred, the average value per payment remains extremely low. This raises questions about whether the activity reflects real demand or just testing or automation.
Wallet Activity Shows Rapid Growth
Wallets funded in the second quarter of 2026 saw a high rate of additional funding within 30 days. That compares to a lower rate for wallets funded in the third quarter of 2025.
The time between receiving funds and making a first x402 payment dropped significantly. This shows the ecosystem is maturing. Developers now build systems that fund, activate, and deploy AI agents with minimal human input.
Limitations and Key Questions
The data does not confirm real commercial demand. Many transactions may be automated or part of test environments. Chainalysis excluded wallets that made only one payment.
China’s activity in security tools doesn’t mean it lacks interest in AI inference. The preference for open-source models may reflect different development choices.
There’s no evidence AI agents are currently generating revenue or fulfilling business needs through these payments. Partnerships with Visa, Mastercard, and Stripe don’t prove real economic value.
What to Watch Next
Future research should track whether AI agent spending grows beyond micropayments. It should assess if these agents are used in real business operations—like data procurement or infrastructure monitoring.
Monitoring AI agents in commercial roles will be key. As more agents are deployed, blockchain’s role in enabling autonomous spending could grow.
For now, AI agent spending is growing—but its economic significance remains uncertain.
Source: Coindesk
For deeper analysis of blockchain-based AI spending, see Ledger’s investigation into crypto theft and UK sanctions on crypto exchanges.
