- BlackRock published a research paper on September 22, 2026, arguing that autonomous AI agents will become a major and underappreciated source of stablecoin demand.
- The firm casts AI as machine-native intelligence and digital assets as machine-native money, with stablecoins positioned to settle high frequency payments between machines.
- BlackRock concedes the market is barely formed: TRM Labs estimated in September that likely AI agents account for just 0.6% to 7.5% of activity on one leading agent payment protocol.
BlackRock, the world’s largest asset manager with more than $10 trillion in assets, said autonomous AI agents could become one of the biggest and most overlooked drivers of stablecoins demand, in a research paper published on September 22, 2026. The BlackRock AI agents thesis carries unusual weight because it comes from the firm that runs the largest bitcoin exchange traded fund.
The 11 page paper, titled “The Machine-Native Economy,” was written by BlackRock’s Digital Assets Research team: Will Su, head of digital assets research; Robert Mitchnick, head of digital assets; Jay Jacobs, U.S. head of equity ETFs; and William Helm, head of U.S. iShares product innovation. The core claim is that AI supplies the intelligence to make decisions, and digital assets supply the money to act on them.
“AI represents machine-native intelligence, while digital assets represent machine-native money,” the authors wrote.
Why AI Agents Need Stablecoins
The near term case rests on payments. As agentic systems, meaning software that can plan and complete multistep tasks with limited human oversight, begin buying data feeds, API access, digital services, and computing power on their own, they need a way to pay that does not wait for a person to click confirm. BlackRock argues that bank accounts and card networks are poorly suited to that job, because account setup and authorization involve humans, and merchant fees make payments worth a fraction of a cent uneconomic.
Stablecoins fit because their value holds steady and blockchains settle around the clock without a bank account. BlackRock put adjusted stablecoin transaction volume above $11 trillion in 2025, in the same range as Visa and Mastercard annual payment volumes, though still far below the $93 trillion that moved through the ACH bank transfer system. Stablecoin volume has grown at roughly 80% a year since 2020, against about 8.5% for ACH, and the circulating supply passed $300 billion in September 2026.
The paper points to infrastructure already live. Coinbase’s x402 protocol, built on an otherwise unused HTTP response code, lets software pay for an API call inside the same request that asks for it, with no account setup. Circle’s Agent Stack gives agents the ability to hold USDC, find services, and pay for them, and Google, Visa, and Stripe have each introduced agent payment systems of their own.
Compute as the Next Asset Class
BlackRock’s second argument is more speculative. It treats computing power as a resource that could be tokenized, traded, and used as collateral. The firm cited consensus estimates that the combined revenue of Amazon, Microsoft, and Google’s cloud units could approach $1.1 trillion by 2030, a compound annual rate near 29%, alongside a Goldman Sachs projection of more than $5 trillion in cumulative AI capital spending between 2025 and 2030. Standardized claims on compute could become “a significant digital asset use case for financing and programmable settlement,” the authors wrote, with products such as compute futures traded on exchanges emerging over time.
To test the thesis, BlackRock ran model simulations rather than observed agent behavior. Bitcoin drew 79.1% of responses in long term store of value scenarios, while stablecoins led in the payment scenarios.
The firm was candid that the machine economy has barely started, describing the ecosystem as nascent with limited real activity and liquidity today. TRM Labs estimated on September 9 that likely AI agent activity accounts for between 0.6% and 7.5% of genuine commerce settled through the x402 protocol. For now, the paper reads as a map of infrastructure to watch rather than proof of a machine economy that has arrived. Whether the agents show up, and when, is the question BlackRock leaves for its institutional audience to weigh.
Editorial Note: Reported and edited by the Crypto India Magazine editorial team. We use AI tools to assist with research and drafting; every article is reviewed and fact-checked by our editors.
Interested in advertising with CIM? Talk to us!
The $4.5 Billion Crypto VCs Raised in A Bear Market Came with A Different Pitch, and the New One Is Spelled “AI”
Crypto startup funding hit a 12-month low in April 2026, yet a16z crypto, Haun Ventures, Dragonfly, and Framework Ventures closed billions in new funds. Polina examines why LPs kept writing checks and how stablecoins, tokenization, and AI agents became the pitch that worked.
Workato Opens Hyderabad AI Hub to Co-Build Agentic AI With Customers
Workato has opened its third AI Collaboration Hub in Hyderabad, after San Francisco and Singapore. The site at Aurobindo Orbit puts customers, GSI partners, and Workato engineers in one room to co-build agentic AI against live enterprise systems. Workato India employs more than 200 people.
Boltz Suspends Bitcoin Swaps as AI-Assisted Attacks Outpace Its Fixes
Boltz has disabled its Bitcoin, Lightning, and Liquid swaps indefinitely after months of automated, AI-assisted probing of its infrastructure. The non-custodial bridge says no user funds were at risk, though attackers now iterate faster than its team can patch.
Clarkson University’s Rex AI Framework Targets Sharper Image Editing and Drug Discovery
Clarkson University researchers built Rex, an AI framework that lets diffusion and flow-matching models reverse their own process with far less error. The work was accepted as an Oral at ICML 2026, the top 0.7% of submissions.