BlackRock Redefines Crypto Money for Machines with Bitcoin and Stablecoins

BlackRock has published a study that redefines the concept of crypto money for machines, arguing that credit cards and bank transfers are no longer sufficient for artificial intelligence agents.

Summary

• Key points
• BlackRock identifies limitations of traditional money for AI agents
• A two-tier system for crypto money for machines
  • Stablecoins as money for expenses
  • Bitcoin as money for savings
• AI models prefer Bitcoin for savings and stablecoins for expenses
• Innovations and developments in the crypto money market for machines
• FAQ
  • Why does BlackRock believe that cards and bank transfers are inadequate for AI agents?
  • What is BlackRock's proposed solution for AI agents' money?
  • How do AI models prefer to save and spend money, according to the cited study?
  • Is the future of machine money exclusively crypto, according to BlackRock?

Key Points

• According to BlackRock, cards and bank transfers fail to meet the economic needs of AI agents.
• The document proposes a two-tier system: stablecoins for expenses, Bitcoin for savings.
• A study cited in the paper shows that 36 AI models chose Bitcoin as a store of value 79.1% of the time.
• The same models chose stablecoins for expenses 53.2% of the time, compared to less than 9% for bank money.
• The circulating liquidity of stablecoins exceeds $300 billion, with transaction volumes surpassing $11 trillion in 2025.

BlackRock Identifies Limitations of Traditional Money for AI Agents

BlackRock's new paper starts from a simple observation: AI agents have payment needs that differ from those of humans, and traditional financial instruments are not designed to serve them. Credit cards and bank transfers, the company writes, cannot handle instantaneous, automated, and continuous transactions between machines.

This is the starting point for a broader thesis: if machines are to exchange value autonomously, a monetary infrastructure specifically designed for them is needed. This is where crypto money for machines comes into play, described as an alternative capable of filling that gap.

A Two-Tier System for Crypto Money for Machines

BlackRock does not propose a single solution but a division of tasks. The document outlines a two-tier system in which stablecoins manage daily spending functions, while Bitcoin, according to the study cited by the company, serves as a store of value.

"Stablecoins, native crypto assets, and other on-chain assets can serve as native tools for machines," BlackRock wrote in the paper.

Stablecoins as Money for Expenses

For frequent and small payments, stablecoins offer the speed and programmability that the traditional banking system does not guarantee. It is the piece designed for continuous operational transactions between AI agents, where each exchange must occur in real-time.

Bitcoin as Money for Savings

For value preservation over time, BlackRock looks to Bitcoin. The logic is that of a scarce asset not subject to discretionary monetary policy decisions, characteristics that make it, according to the paper, suitable for the role of reserve for machines.

AI Models Prefer Bitcoin for Savings and Stablecoins for Expenses

The most cited part of the paper does not originate from BlackRock but from an external study that the company presents as "preliminary support" for its thesis. The study, conducted by the Bitcoin Policy Institute, a pro-Bitcoin research group, tested 36 cutting-edge AI models created by Anthropic, OpenAI, Google, xAI, and DeepSeek, gathering 9,072 responses.

Models were asked where to store value, and the predominant answer was Bitcoin, chosen 79.1% of the time. When asked what to use for spending, models indicated stablecoins in 53.2% of cases. Traditional bank money was chosen less than 9% of the time in both scenarios.

At the time of publication, Bitcoin was trading near $86,400, up about 0.6% in the previous 24 hours, according to market data reported by BeInCrypto. It should be noted that every sentence with future projections in the report uses verbs like "could" or "may," and a note clarifies that this is not a financial forecast.

Innovations and Developments in the Crypto Money Market for Machines

BlackRock does not describe an exclusively crypto future for machine money. The paper lists alternative payment rails proposed by Stripe, OpenAI, Google, and Visa, some of which still regulate transactions in traditional bank money. The company also acknowledges that the actual payment volumes between AI agents remain small today.

Among the cited examples is the x402 protocol from Coinbase, which revives the HTTP 402 "Payment Required" code, dormant for years, allowing agents to pay instantly for access to data or services.

The scale achieved by stablecoins gives concrete weight to BlackRock's thesis. The company estimates circulating liquidity exceeding $300 billion, with adequate transaction volume surpassing $11 trillion in 2025, a level close to Visa's $11.2 trillion during the same period, although still distant from the $16.7 trillion overall moved by the payment giant. In the first half of 2026, stablecoins moved another $8.5 trillion.

BlackRock, after all, does not observe this market from the outside. Before this year, the company launched a money market fund designed for stablecoin issuers, which they can use to park their reserves. This detail clarifies why the company looks with interest at a sector where it is already present as a provider of financial infrastructure.

FAQ

Why does BlackRock believe that cards and bank transfers are inadequate for AI agents?

BlackRock's new paper states that credit cards and bank transfers cannot meet the specific monetary needs of artificial intelligence agents.

What solution does BlackRock propose for AI agents' money?

BlackRock proposes a two-tier system, with stablecoins used for spending and Bitcoin used as a savings reserve by machines.

How do AI models prefer to store and spend money, according to the cited study?

36 AI models chose Bitcoin as a store of value about 79.1% of the time and stablecoins for spending about 53.2% of the time.

Is the future of machine money exclusively crypto, according to BlackRock?

No, the paper does not describe an exclusively crypto future and acknowledges the existence of other payment rails, in addition to current volumes still small for AI agents.

Content created with the assistance of artificial intelligence and human editorial review.