Over the past year, blockchain networks have seen a remarkable surge in payments made by artificial intelligence agents. Data reveals that these autonomous software agents processed 176 million transactions totaling $73 million. The average amount transferred per transaction was only $0.31.
A new era for blockchain micro-payments
Autonomous software agents are applications that independently make purchases and payment decisions on behalf of users. Recently, these agents have started using blockchain-based micro-payments to cover necessities like API access, data streams, and cloud server rentals. Transactions averaging just $0.31 are considered uneconomical using traditional payment systems. For example, payment giants such as Visa charge a fixed $0.30 fee per transaction, exposing significant inefficiencies with small-value payments.
As demand from AI agents for low-value, high-frequency transactions grows, it pushes beyond the limits of traditional financial systems, while blockchain infrastructures can handle this demand at much lower costs.
The advent of blockchain networks such as Base and Tempo, combined with stablecoin-based payments, has dramatically reduced transaction fees—sometimes to less than a cent. Latest figures show more than 104,000 AI agents are active and listed across over 15 directories. Most of these agents routinely and automatically pay for data feeds, cloud resources, or AI services. All payments are executed instantly and autonomously, without any human approval.
Glossary: An autonomous AI agent is software that can independently perform transactions for a user. These programs manage tasks such as purchasing, payments, or data access automatically, without human intervention.
Tech giants move to capture the autonomous payments market
The high potential of this market has prompted major technology companies to develop new payment solutions. Coinbase recently launched the x402 protocol, enabling service payments directly with USDC—no membership or account required. Stripe partnered with the Tempo blockchain to develop the Machine Payments Protocol. Google has introduced AP2, a system that authorizes autonomous agents to make purchases. Meanwhile, Visa is focusing on token-based network infrastructures specifically tailored for AI applications.
According to analysis by Keyrock, large financial firms have invested over $8 billion in acquisitions to secure a competitive edge. These investments are laying the groundwork for systems that could become the backbone of the future autonomous AI economy.
Heavy reliance on USDC brings systemic risk
Recent data indicates that an overwhelming 98.6% of all AI agent payments were carried out with USDC, a stablecoin issued by Circle. Such strong reliance on a single digital asset raises both Circle’s market profile and systemic risk concerns. Should Circle face regulatory intervention, devaluation, or prolonged outages, nearly the entire agent economy would lack viable alternatives.
| Payment Method | Market Share (%) |
|---|---|
| USDC | 98.6 |
| Other | 1.4 |
While regulatory frameworks such as MiCA in Europe, the GENIUS Act in the United States, and the EU’s new AI Law are expected to take effect in mid-2026, there is currently no direct legal framework for autonomous payments. Uncertainties still surround areas like machine-to-machine commerce, identity verification, and liability.
Market outlook and future prospects
Despite rapid growth, the current transaction volume remains modest compared to established financial players—for instance, Visa processes $14.5 trillion in payments annually. Nonetheless, analysts and research firms expect that AI-powered payments will expand into a much larger market over the coming years. Gartner projects that autonomous agents could process $15 trillion in payments annually by 2028. McKinsey anticipates that AI-driven retail commerce may reach $3 trillion to $5 trillion by 2030. The necessary infrastructure is being built now, with significantly higher transaction volumes expected in the medium term.
