BlackRock Says AI Agents Will Power a Machine-Native Economy Built on Stablecoins

BlackRock published a research paper in September 2026 titled "The Machine-Native Economy", arguing that widespread AI adoption represents an underappreciated source of demand for digital assets — particularly stablecoins, native cryptocurrencies, and tokenized real-world assets.

The world's largest asset manager frames the thesis as structural adoption logic, not a near-term price prediction.

Machine-to-machine payments

Authors Will Su, Robert Mitchnick, Jay Jacobs, and William Helm contend that AI agents conducting commerce on behalf of users and businesses will require payment infrastructure optimized for:

• High-frequency transactions
• Sub-cent micropayments
• 24/7 settlement without banking hours

Traditional payment rails struggle with these requirements. BlackRock argues programmable on-chain assets — especially stablecoins for spending — are better suited to agent commerce.

The paper cites Coinbase's x402 protocol and Tempo's Machine Payments as early infrastructure examples. Block's September 25 announcement adding Bitcoin Lightning to x402 adds another data point.

Bitcoin as store of value, stablecoins as spend

BlackRock references a February 2026 Bitcoin Policy Institute study testing 36 frontier AI models across 9,072 responses. Models selected Bitcoin as store of value 79.1% of the time and stablecoins for spending 53.2% of the time.

That behavioral split — if replicated in production agent wallets — has direct implications for treasury strategy, float management, and payment vendor selection.

Tokenized compute as a second market

Beyond payments, BlackRock identifies tokenized compute — representing claims on AI processing capacity — as a distinct opportunity. AI companies may want to lock in compute costs; providers may want to hedge demand volatility. Tokenized claims could be transferred, collateralized, or traded, broadening institutional participation.

Business implications for founders

Fintech and crypto infrastructure startups serving agent payments may find institutional validation accelerating enterprise sales cycles.

AI application companies should model payment flows explicitly — agents that cannot pay autonomously become bottlenecks in multi-step workflows.

Enterprise finance teams should prepare for machine-initiated spend with policy engines, approval gates, and accounting categories that did not exist five years ago.

Skepticism worth holding

BlackRock has commercial interests in digital asset products. The machine-native economy thesis is directionally compelling but unproven at scale — agent commerce remains early, and regulatory frameworks for autonomous spending are immature.

Still, when the world's largest asset manager publishes structural adoption logic for stablecoins tied to AI agents, venture investors and corporate strategists pay attention — because capital allocation follows credible institutional narratives, even before full market proof.