BlackRock: AI Agents Could Drive the Next Stablecoin Boom
BlackRock's 'Machine-Native Economy' paper argues AI agents will pay each other in stablecoins, and that most AI models it tested favour Bitcoin for saving.
The world's largest asset manager thinks the next big buyers of crypto may not be people at all. In a research paper titled "The Machine-Native Economy", published on 22 September 2026, BlackRock's digital assets research team argues that AI agents, software that acts and spends on its own, could become a major new source of demand for stablecoins and other digital assets.
Key takeaways
- The paper comes from BlackRock's Digital Assets Research team, led by Will Su and Robert Mitchnick.
- Its core claim: stablecoins suit the tiny, round-the-clock payments AI agents will need better than banks or card networks.
- Stablecoins are worth over $300 billion, and adjusted transaction volume reached about $11.2 trillion in 2025.
- In simulated tests, AI models chose Bitcoin as a long-term store of value 79.1% of the time.
- BlackRock admits agent payments are still "nascent".
The argument
AI agents are programs that can carry out tasks on their own: booking services, buying data, paying for computing power, or calling other agents. To do that, they need a way to pay. BlackRock's case is that traditional payment systems were built for humans. They close on weekends, charge fixed fees that make very small payments pointless, and need accounts tied to a person or a company.
Stablecoins avoid most of that. They run 24 hours a day, can settle payments of less than a cent, and can be held by any wallet, including one controlled by software. That makes them, in BlackRock's view, the natural money for machine-to-machine payments.
The numbers behind it
The paper puts the stablecoin market at more than $300 billion as of September 2026. Adjusted stablecoin transaction volume was about $11.2 trillion in 2025, in the same range as Visa and Mastercard's yearly payment volumes, though BlackRock notes these figures are not directly comparable. It is still far below the roughly $93 trillion that moved over the US ACH bank transfer system that year. But stablecoin volume has grown about 80% a year since 2020, against roughly 8.5% for ACH.
What the AI models preferred
BlackRock also tested 36 AI models across 9,072 scenarios to see which assets they would choose for different jobs. For long-term saving, the models picked Bitcoin 79.1% of the time. The paper is careful to say these are simulated answers, not what real agents actually do with real money.
It also points to computing power as a possible new asset class, with cloud revenue projected at about $1.1 trillion by 2030, and suggests products such as exchange-traded compute futures could follow.
The caveats
BlackRock does not claim this is happening at scale yet. It calls agent payment activity nascent and says liquidity in compute markets is limited. Many practical problems, from how to settle payments for computing to energy costs, remain unsolved. Treat the paper as a direction of travel from a very influential firm, not a forecast with dates attached.
Why it matters for Indian readers
When BlackRock publishes research like this, it tends to shape how big investors think about the sector. BlackRock already runs the largest spot Bitcoin ETF and a tokenized money market fund. For Indian users, stablecoins such as USDT are mostly a way to move between trades, and they carry the same 30% tax and 1% TDS as other crypto. One USDT is worth about ₹95.96 right now; see the Tether price in INR, or read what a stablecoin is if you are new to them.
FAQ
What is BlackRock's "Machine-Native Economy" paper?
A research paper published on 22 September 2026 by BlackRock's digital assets team. It argues that AI agents could become a big source of demand for stablecoins and other digital assets.
Why would AI agents use stablecoins?
Stablecoins work around the clock, can handle payments of under a cent, and can be held by software wallets, which suits automated payments between machines.
Did BlackRock say AI prefers Bitcoin?
In simulated tests of 36 AI models, Bitcoin was chosen as a long-term store of value 79.1% of the time. BlackRock stresses these were simulations, not real agent behaviour.
This article is AI-assisted, educational and general in nature. It is not financial advice and never a guarantee of profit. Every trade is at your own risk on your own exchange. See our risk disclosure and editorial policy.
