Artificial intelligence agents are rapidly becoming active participants in the crypto economy, with a new report revealing that machine-to-machine payments have evolved from a simple concept into a growing financial ecosystem over the past year.
According to a report released Thursday by crypto investment firm Keyrock in partnership with Coinbase and blockchain company Tempo, AI agents processed more than $73 million in settlements across roughly 176 million transactions between May 2025 and April 2026.
The report highlights how stablecoins and blockchain-based payment infrastructure are increasingly becoming the default financial rails for AI-driven commerce.
AI Agents Moving From Experiment To Real Economic Activity
Keyrock researcher Ben Harvey stated that AI agent payments have quickly matured over the past 12 months, transforming from experimental technology into a functioning ecosystem with real transaction volume and infrastructure investment.
Harvey noted that major industry players have already deployed more than $8 billion through acquisitions and strategic investments as companies compete to secure positions within what is emerging as an entirely new digital payment layer.
AI agents are becoming increasingly important within crypto and Web3 ecosystems. Many industry leaders believe autonomous agents capable of sending payments, executing trades, interacting with protocols, and managing digital services could eventually become one of the largest drivers of blockchain adoption.
Earlier this year, Circle CEO Jeremy Allaire predicted that billions of AI agents could be using stablecoins on behalf of users within the next five years.
Traditional Payment Systems Struggle With AI Commerce
One of the report’s most important findings is how traditional payment infrastructure remains poorly suited for machine-driven microtransactions.
By the end of the first quarter of 2026, more than 104,000 AI agents had reportedly been registered across at least 15 different directories and registries.
The average transaction size among these agents was only around 31 cents.
According to Harvey, that transaction size explains why conventional payment systems are economically inefficient for AI-driven commerce.
Traditional payment processors often charge fixed fees near 30 cents per transaction, making extremely small payments impractical.
For example, an AI agent paying only a few cents to access a weather-data API or complete a small digital task cannot efficiently use legacy payment systems like Visa or traditional banking rails.
Stablecoins Solve The Microtransaction Problem
The report argues that stablecoins effectively became the dominant settlement layer for AI agents almost by default because blockchain payment rails can process very small transactions efficiently without collapsing under fixed fee structures.
This capability allows AI agents to autonomously:
- Pay for digital services
- Access APIs
- Execute Web3 functions
- Trade assets
- Launch tokens
- Interact with decentralized protocols
AI agents are also increasingly being explored for automated trading and portfolio management within the crypto industry.
A CoinGecko survey conducted last year found that 87% of crypto users would allow AI agents to manage at least 10% of their crypto portfolios.
USDC Dominates AI Agent Settlements
The report also revealed that more than 98% of AI-agent settlements were conducted using Circle’s USDC stablecoin.
Harvey described this dominance as both a major validation for USDC and a potential systemic vulnerability for the broader AI-agent economy.
Because nearly the entire ecosystem depends heavily on a single stablecoin issuer, the report warns that major operational or regulatory issues affecting Circle could create significant risks across the sector.
According to Harvey, challenges such as:
- Regulatory pressure
- Stablecoin depegging
- Infrastructure outages
- Reserve management concerns
could severely disrupt AI-agent payment systems if alternative settlement options remain underdeveloped.
The report argues that this concentration risk is not currently receiving enough attention despite the ecosystem’s rapid growth.
A New Financial Layer Emerging
The rise of AI agents using crypto infrastructure represents a major shift in how digital commerce may operate in the future.
Rather than humans initiating every online payment or transaction manually, autonomous software agents could increasingly handle:
- Payments
- Subscriptions
- Trading
- Service access
- Data purchases
- Financial operations
directly on behalf of users.
As blockchain infrastructure, stablecoins, and AI systems continue evolving together, many analysts believe the combination could create an entirely new internet-native financial layer optimized for machine-driven economic activity.
While the ecosystem remains in its early stages, the rapid growth in transaction volume and infrastructure investment suggests that crypto-powered AI commerce may already be moving beyond experimentation and into real-world adoption.
