BlackRock Predicts AI Agents Will Need Stablecoins for Machine-to-Machine Payments

BlackRock's Digital Assets Research team has published a report titled 'The Machine-Native Economy' which explores the potential for AI agents to transact in cryptocurrency. The report argues that AI gives machines native intelligence, while digital assets provide machine-native money, and the two are built for each other.

The paper notes that current payment systems, such as credit cards and bank wires, are not suitable for AI-to-AI transactions, as they require human approval and have high fees. Stablecoins, on the other hand, can settle around the clock without a bank teller or card network taking a cut.

BlackRock estimates that the circulating stablecoin market cap is over $300 billion, with adjusted transaction volume topping $11 trillion in 2025. This growth rate of 80% compound annual since 2020 far exceeds that of ACH, which grew at around 8.5% a year.

The report specifically names Ethereum and Circle's Arc as the settlement venues most likely to carry AI-to-AI traffic, with Coinbase's x402 protocol emerging as a standard for machine-to-machine payments. BlackRock also projects combined AWS, Azure, and Google Cloud revenue reaching $1.1 trillion by 2030, and suggests that standardized claims on compute capacity could become tradable collateral.