BlackRock Says AI Agents Could Turn Blockchains Into Machine-Native Settlement Networks
BlackRock is highlighting a growing connection between artificial intelligence and digital assets, arguing that increasingly autonomous AI systems could create new demand for programmable financial infrastructure. In its latest research, the asset manager describes AI as “machine-native intelligence” and digital assets as “machine-native money,” with blockchains potentially connecting the two.
The research paper, titled “The Machine-Native Economy,” identifies three major areas where AI and digital assets could converge: machine-to-machine payments, programmable tokenized financial assets and the emerging market for computer resources. BlackRock argues that these applications could expand the utility of blockchain infrastructure beyond its historical role in cryptocurrency trading.
AI Agents Need Machine-Native Payments
BlackRock points to agentic AI as an important driver of this change. Unlike conventional software that waits for instructions, AI agents can plan and execute multistep tasks while interacting with external tools and services. That creates a need for payment systems capable of handling transactions automatically, continuously and at very small values.
The research highlights stablecoins as one potential settlement instrument for this economy. Protocols including Coinbase’s x402, Stripe and Tempo’s Machine Payments Protocol, and Stripe and OpenAI’s Agentic Commerce Protocol are examples of emerging systems designed to connect automated software activity with payments. BlackRock notes that traditional payment networks will continue to matter, particularly when agents interact with human-operated businesses and consumers.
The scale of stablecoin activity is already providing a significant base for this thesis. BlackRock’s research says adjusted stablecoin transaction volume exceeded $11 trillion in 2025, although it cautions that the measurement is not directly comparable with traditional payment-network volumes because different methodologies and transaction categories are involved.
BlackRock also highlights the rapid growth of the stablecoin market itself, which it says exceeded $300 billion in circulating market capitalization by September 2026. The firm argues that improving regulatory frameworks in the United States, Europe and Asia-Pacific could provide additional support for stablecoin adoption.
For blockchain networks, increased machine-to-machine activity could create demand for blockspace, validators and transaction infrastructure. The impact on individual native tokens would depend on network economics, including transaction fees, staking mechanisms and whether applications or users can sponsor gas costs.
The second major area is tokenized financial assets. BlackRock argues that AI agents could eventually evaluate balances, rules and eligibility before executing transactions involving tokenized funds, securities and other real-world assets through smart contracts. Such systems could reduce the need for bespoke integrations by giving agents standardized, machine-readable representations of ownership and transaction rules.
Compute Could Become a New Digital Asset Market
BlackRock also identifies computing capacity as a potential new market for digital assets. As AI models require increasing amounts of processing power, access to GPUs, specialized hardware and data-center capacity could increasingly become an economic resource that can be priced, financed and traded.
The paper cites consensus estimates that combined revenue from major hyperscaler cloud businesses could reach approximately $1.1 trillion by 2030. BlackRock argues that standardized claims on compute capacity could eventually be represented digitally, transferred between parties, pledged as collateral and settled through programmable infrastructure.
That market remains considerably less developed than stablecoins or tokenized financial assets. Differences in chip generations, regional electricity costs, hardware performance and delivery requirements create substantial challenges for designing standardized compute contracts. BlackRock nevertheless sees potential for futures, contracts for difference and other market structures to improve price discovery and hedging.
AI agents could eventually use these markets autonomously. An agent running a demanding task could compare available computing capacity based on price, performance, latency, location and hardware, then provision resources and settle payment on a usage basis. Protocols such as x402 could potentially support payment per API call, model token, compute job or other unit of consumption.
BlackRock’s conclusion is not that this machine economy already exists at scale. The firm explicitly describes agentic payment activity and compute-market liquidity as nascent. Instead, its thesis is that as AI systems become more capable and take on a larger role in economic activity, digital assets could become increasingly useful as the financial infrastructure connecting machines, assets and resources.
The shift would represent a significant change in the role of blockchain networks. Rather than serving primarily as markets for people trading digital assets, they could increasingly function as programmable settlement infrastructure for autonomous software. Whether that becomes a major source of blockchain demand will depend on adoption, regulation, network economics and the ability of these systems to operate reliably at machine scale.
About The Author
