BlackRock, in its research The Machine-Native Economy, stated that the widespread adoption of AI and the growth of machine-to-machine payments could significantly increase demand for blockchains, stablecoins, and other programmable payment rails. The company also sees potential for digital assets in the computing power market: claims to access computing resources could be tokenized, traded, and used as collateral.

Agent-Based AI Increases the Need for Machine-Native Money

According to BlackRock, agent-based AI will require tools for round-the-clock, high-frequency micropayments. Traditional payment infrastructure often requires human involvement at account opening and authorization stages, while fees make small transactions unprofitable and settlement times vary. Stablecoins, native cryptocurrencies, and tokenized real-world assets are better suited for sub-cent M2M payments, with stablecoins expected to lead transactional use.

Tokenization of Computing Power as a New Market

Demand for AI computing is rising, and companies may seek to lock in prices and suppliers. Claims to computing power, represented as tokens, could be transferred, pledged, and traded, potentially expanding institutional investor participation. Agent-based AI would be able to automatically acquire necessary resources in such markets.

Industry Already Developing Payment Protocols for AI

The industry is rolling out solutions for automated payments by AI agents: Coinbase has introduced the x402 protocol, and Tempo has launched the Machine Payments Protocol. In May, Circle announced agent wallets and USDC payment tools, while OKX developed the Agent Payments Protocol for recurring debits and escrow scenarios.