AI agents settled $73 million in crypto payments over 12 months
Artificial intelligence agents have moved from experimental concept to active payment infrastructure, settling $73 million across 176 million transactions between May 2025 and April 2026, according to a report from crypto investment firm Keyrock released Thursday.
The data comes from collaboration with crypto exchange Coinbase and blockchain platform Tempo. Over the same period, major tech companies deployed more than $8 billion in acquisitions to establish footholds in what researchers describe as an emerging payment stack for machine-to-machine commerce.
Why traditional payment systems don't work for AI agents
The average transaction size was 31 cents. That figure exposes a fundamental problem: conventional payment processors charge roughly 30 cents per transaction, making sub-dollar payments economically unviable.
An AI agent calling a weather API for 3 cents cannot route through Visa or traditional networks without losing money on fees. Stablecoins-crypto tokens pegged to fiat currency-became the default settlement layer because they were the only instrument that could handle these small transactions without collapsing the economics.
By the end of Q1 2026, more than 104,000 agents were registered across 15 or more directories and registries.
One stablecoin dominates the ecosystem
Circle's USDC handled more than 98% of all AI agent settlements. This concentration creates both validation and risk.
The entire ecosystem depends on a single stablecoin issuer's reserve management, regulatory standing, and technical infrastructure. If Circle faces a regulatory challenge, a price de-peg, or sustained downtime, the agent economy has no fallback option.
This systemic dependency has received little public scrutiny despite the potential impact as transaction volumes increase.
Broader adoption signals growing comfort with autonomous systems
AI agents are already being deployed for Web3 applications, token launches, and autonomous protocol interaction. Some platforms are exploring AI for trading and portfolio management.
A 2025 survey of 2,632 crypto users found that 87% said they would allow AI agents to manage at least 10% of their crypto portfolio, indicating significant user comfort with autonomous financial systems.
For IT and development professionals, understanding how AI systems integrate with payment infrastructure has become relevant to broader infrastructure decisions. The technical requirements for autonomous AI agents operating at scale differ fundamentally from traditional application architecture.
Your membership also unlocks:
