BlackRock says AI agents could drive fresh demand for stablecoins and blockchains, and sees tokenized computing power opening a new market as cloud revenue heads toward $1.1 trillion by 2030.
Key Points:
- BlackRock says AI agents will need machine-native payment rails, with stablecoins likely to lead.
- The firm sees claims on AI computing capacity becoming tradable digital assets.
- The authors warn agentic payments and compute markets are still at an early stage.
BlackRock Stablecoin Thesis
The world's largest asset manager laid out the case in a paper titled "The Machine-Native Economy," written by Will Su, Robert Mitchnick, Jay Jacobs and William Helm. The authors argued that AI agents, which plan and carry out tasks with little human input, will need payment tools built for machines rather than people. They expect stablecoins to lead.
Card networks and bank transfers already handle plenty of automation, the paper said, but opening an account can still require a person and merchant fees make tiny payments uneconomic. Card settlement and dispute finality can take longer still.
Stablecoins in circulation topped $300 billion as of September, according to the report. Adjusted stablecoin volume passed $11 trillion in 2025, broadly in line with the annual volumes of Visa and Mastercard. That still trailed the $93 trillion moved over the U.S. ACH network last year, though stablecoin volume grew about 80% a year from 2020 to 2025, against roughly 8.5% for ACH.
Also Read: Dogecoin Outruns Every Major Token As Bears Get Squeezed Out
Tokenized AI Compute Market
The authors see a second opening in compute, the processing power used to train and run AI systems. Analyst estimates cited in the paper put combined 2030 revenue for Amazon Web Services, Microsoft's Intelligent Cloud unit and Google Cloud at about $1.1 trillion, a 29% annual growth rate from 2025.
Standardized claims on that capacity could be transferred, pledged as collateral and settled on-chain, BlackRock said. AI agents could then shop for compute by price, latency and location, and the paper cites Stripe's August deal to buy OpenRouter as an early sign of that shift. Native tokens may gain too, since heavier traffic on networks such as Ethereum (ETH) could lift demand for blockspace.
The authors called AI "a structural catalyst for digital asset adoption" and said investors still underrate the link. They also cautioned that agentic payment activity and compute-market liquidity remain limited for now.
Brian Armstrong AI Crypto Case
Crypto executives have pushed the same idea for months. Coinbase CEO Brian Armstrong wrote on Jul. 26 that AI makes crypto more important because agents cannot open bank accounts or wait days for a wire. He credited Coinbase's x402 protocol, its Base network and USD Coin (USDC) with powering most agentic payments, and BlackRock's paper now flags x402 as a possible standard for fast machine payments.
Read Next: OpenAI Faces British Columbia Lawsuit Over 8-Victim School Shooting
