The payment ecosystem for artificial intelligence is gradually taking clear shape, with room for both traditional cards and stablecoins. My analysis of the latest data shows that the division of functionality between these instruments is becoming inevitable and logical.
Experts identify two fundamentally different types of transactions conducted by AI agents. The first is purchases made on direct human instruction: booking tickets, managing subscriptions, ordering goods. Here, cards remain the optimal solution, as the amounts are comparable to regular retail trade.
The second type is micropayments between programs, where amounts regularly drop below $1. This involves paying for individual API requests, accessing computing power, or purchasing data. It is here that traditional card infrastructure fails due to fixed fees that eat up the entire economics of the transaction.
Why stablecoins win in micropayments
On the Base and Solana networks, transaction fees amount to fractions of a cent. This makes blockchains the only viable option for frequent small transfers between AI agents. The report examines two open protocols: x402, active on Base, Solana, and Polygon, as well as the Machine Payment Protocol (MPP), which supports settlements both through blockchains and traditional systems.
The real data is impressive. As of April 21, after filtering out fake and test operations, approximately $15 million passed through x402 across 109.6 million transactions. MPP's figures are more modest — roughly $25,000 and 115,000 transfers. It is important to note that raw data was significantly higher: $135.7 million and 178.3 million operations via x402, but a substantial portion was discarded as irrelevant activity.
The main stumbling block — trust
The key unresolved issue is liability. Who will be responsible for an erroneous or unauthorized purchase: the AI agent owner, the platform operator, the model developer, or the seller? Existing chargeback mechanisms are designed for human operations, not for thousands of automatic transactions per hour. Additional risk comes from malicious instructions capable of redirecting payments.
In my view, we are witnessing the emergence of a hybrid model, where cards will remain for authorization and purchases on human instruction, while stablecoins will handle settlements between programs. This is not competition, but a division of labor. In June, Alchemy and Visa already launched the AgentCard service for purchases through AI agents within user limits — the first step toward formalizing this new payment reality.
Expert opinion: The payment market for AI agents is in its infancy, but it is already clear that stablecoins will occupy a niche inaccessible to traditional cards. The issue of trust and liability will become the main driver of regulation in the next 12–18 months. Those who solve it first will gain a competitive advantage.