BlackRock (NYSE: BLK) has published a new research paper examining how artificial intelligence and digital assets are starting to reinforce one another. Titled, The Machine-Native Economy, the note from the firm’s digital-assets team argues that widespread AI adoption could become an underappreciated source of demand for blockchains, stablecoins, and tokenized instruments.
The authors frame the two technologies as complementary: AI supplies machine-native intelligence, while digital assets supply machine-native money.
That pairing becomes more important as “agentic” systems—software that can plan and complete multi-step tasks with limited human input—begin buying data, booking services, and renting processing power on their own.
Large language models and blockchains already share a similar architecture: both convert real-world inputs into tokens machines can process.
The paper suggests this common design could give autonomous agents a more direct interface with programmable financial infrastructure.
Payments sit at the center of the near-term thesis.
Traditional rails such as ACH and card networks handle large volumes but are less suited to always-on, sub-cent, high-frequency transfers. Stablecoins and emerging agent protocols, including Coinbase’s x402, Google’s AP2, and Visa’s TAP, are presented as better matches.
Circulating stablecoin value exceeded $300 billion in September 2026, with adjusted transaction volume above $11 trillion in 2025.
That activity still trails ACH’s $93 trillion, yet stablecoin volumes grew at an 80 percent compound annual rate from 2020 to 2025, versus roughly 8.5 percent for ACH.
A longer-horizon idea is treating compute itself as a digital asset. Hyperscaler cloud revenues from Amazon, Microsoft, and Google could approach $1.1 trillion a year by 2030.
Standardized, tokenized claims on processing capacity could let agents source, finance, and settle compute programmatically.
The paper stresses the market remains early; independent estimates put current AI-agent activity on x402 at only a small slice of volume.
Consulting firms have been mapping adjacent pieces of the same picture.
KPMG has highlighted x402-style standards as a way to support machine-initiated micropayments that existing rails struggle to deliver.
Accenture has focused on “AI tokenomics,” warning that unmanaged token spend can balloon enterprise costs without new governance.
Deloitte projects that agentic systems could influence trillions of dollars in commerce by 2030 and sees stablecoins taking a larger role in retail settlement.
EY’s institutional survey found rising interest in tokenization and blockchain rails for trading and settlement.
PwC has expanded crypto and stablecoin work as regulation clarifies.
Citigroup forecasts a base-case tokenized-asset market of about $5.5 trillion by 2030, with stablecoins helping complete instant settlement.
Oliver Wyman estimates digital rails could reshape wholesale-banking economics and shift tens of billions in market-infrastructure revenue onto tokenized systems, while noting that a liquid compute market still needs standardization and forward curves.
Major financial institutions are already building related capabilities. JPMorgan’s Kinexys platform has processed trillions in blockchain transactions and now supports tokenized money-market funds.
Bank of America has appointed dedicated digital assets and AI leaders and is participating in bank-led tokenized-deposit and stablecoin initiatives.
BNY is expanding crypto custody, tokenized fund infrastructure, and on-premise AI compute while arguing that large banks will serve as the bridge between traditional and digital finance.
Blackstone has poured capital into AI infrastructure, data centers, and power projects, including a large chip-financing vehicle, underscoring how physical compute scarcity sits behind the software story.
The BlackRock research paper as well as these parallel views point to the same conclusion: as agents move from generating text to executing economic actions, programmable money and tokenized resources look less like optional experiments and more like plumbing for an increasingly automated economy. The infrastructure is still nascent, but the direction of travel is becoming clearer.