Summary
- BlackRock's recent report claims stablecoins are more adept than banks or card networks for facilitating sub-cent, 24/7 payments that AI agents will require.
- The document introduces a novel asset category: tokenized claims on computing resources.
- A study by TRM Labs indicates that AI agents currently represent only 0.6% to 7.5% of payment transactions on Coinbase's x402 protocol.
BlackRock anticipates that artificial intelligence will significantly influence the cryptocurrency market's future demand.
The leading asset management firm released a research paper this week asserting that AI, rather than new regulations or institutional investments, might become a pivotal but often overlooked factor in driving crypto demand. According to BlackRock, AI agents are poised to start making purchases independently, with stablecoins being the most suitable payment option.
“As agents become more capable and persistent, standardized claims on compute capacity could become a significant digital asset use case for financing and programmable settlement,” stated BlackRock.
The report, titled "The Machine-Native Economy," is authored by BlackRock's Digital Assets Research team, under the leadership of Will Su and Robert Mitchnick, along with the firm's U.S. equity ETF and iShares product teams.
The primary concept revolves around a clever play on words. “AI represents machine-native intelligence, while digital assets represent machine-native money,” BlackRock explains, emphasizing that these two domains were always destined to converge.
AI Agents Making Purchases
The report focuses on "agentic AI"—AI systems capable of independently planning and executing complex tasks, utilizing external tools and services with minimal human intervention. These agents can do more than just recommend travel; they can book flights, process payments, and provide receipts.
This level of autonomy presents a challenge: a payment mechanism is required.
Conventional payment systems are not designed for this scenario. Opening a bank account or obtaining a credit card necessitates human identification, and card fees are impractical for tiny transactions, such as a fraction-of-a-cent API call.
BlackRock proposes stablecoins as a solution—cryptocurrencies tied to stable assets like the U.S. dollar, ensuring price stability rather than volatility seen in Bitcoin. Stablecoins can operate continuously, settle transactions almost instantaneously, and do not require bank intermediaries.
According to the report, adjusted stablecoin transaction volumes are projected to exceed $11 trillion by 2025, aligning with the annual payment volumes of Visa and Mastercard. While this figure is still overshadowed by the $93 trillion processed through traditional ACH bank transfers that year, stablecoin transaction volume has surged approximately 80% annually since 2020, compared to merely 8.5% for ACH.
“The rise of agentic AI and machine-to-machine payments will likely increase demand for blockchains and other programmable payment infrastructure; stablecoins, native cryptoassets, and other on-chain assets can serve as machine-native instruments for payment and settlement across these rail,” the report asserts.
Existing Infrastructure
Some of the necessary infrastructure is already operational. The x402 protocol, created by Coinbase, utilizes the rarely used "HTTP 402: Payment Required" web code, allowing software to pay for data feeds or API calls in the same request without needing an account or human approval.
Amazon has integrated stablecoin payments into its AI cloud services via partnerships with Coinbase and Stripe, enabling agents to automatically pay for APIs, data feeds, and bookings during tasks. Google has developed its own payment layer for agents, supported by Coinbase and the Ethereum Foundation, expanding its Agent2Agent framework to manage card, stablecoin, and real-time bank transfer payments.
However, real-world application remains limited. Blockchain analytics company TRM Labs analyzed $52.7 million in x402 transactions this year and estimated that AI agents accounted for only 0.6% to 7.5% of this value, with much of the activity resembling standard automated scripts rather than true autonomous agents.
BlackRock does not claim that widespread adoption will occur immediately. Instead, it believes the necessary infrastructure is being established in anticipation of future demand.
Operating AI systems is costly, and the demand for cloud computing resources is on the rise. BlackRock references analyst predictions estimating that the combined cloud revenues of Amazon, Microsoft, and Google could reach approximately $1.1 trillion by 2030.
The report envisions standardized contracts for computing resources, akin to today's oil or wheat futures, which could be bought, sold, utilized as collateral, or settled automatically on a blockchain. In this scenario, an AI agent could compare prices for server capacity and pay directly for services as needed, without human negotiation.
As of September 2026, BlackRock estimates the market cap for circulating stablecoins to exceed $300 billion, with more than $11 trillion in adjusted transaction volume recorded in 2025 alone. The firm predicts that AI agents, rather than human consumers, will spearhead the next phase of growth as these autonomous systems begin to manage their own expenses.