Key Highlights
- BlackRock examines how AI agents could create demand for automated payment infrastructure.
- Stablecoins could provide a settlement mechanism for machine-to-machine payments.
- Tokenized assets could allow AI systems to interact with financial products through programmable infrastructure.
BlackRock Digital Assets Research is examining how artificial intelligence could interact with blockchain-based financial infrastructure, focusing on automated payments, tokenized assets and computing markets.
In an announcement on September 22, BlackRock outlined its latest research on the emerging relationship between AI and digital assets. In its paper, “The Machine-Native Economy: How digital assets connect intelligence, commerce, and compute,” the firm looks at how increasingly autonomous AI systems could interact with financial infrastructure without requiring human involvement at every step.
The research focuses on three areas: payments between AI agents, access to tokenized financial assets and markets for computing capacity.
AI agents could drive automated payments
As AI systems become capable of completing tasks independently, they may need to pay for data, software, APIs and computing resources.
BlackRock points to emerging payment initiatives including x402, the Machine Payments Protocol (MPP) developed by Stripe and Tempo, and the Agentic Commerce Protocol (ACP) developed by Stripe and OpenAI.
These systems are designed to allow software agents to initiate and settle payments without requiring a person to approve every transaction.
Such infrastructure could be relevant to frequent, low-value transactions between automated systems, although widespread adoption would depend on issues including payment permissions, identity and compliance.
Stablecoins could be used for machine payments
The research also examines stablecoins as a possible settlement mechanism for automated transactions.
Stablecoins can provide blockchain-based settlement while maintaining a value typically pegged to a fiat currency such as the U.S. dollar.
BlackRock cites more than $11 trillion in adjusted stablecoin transaction volume during 2025, while noting that the methodology is not directly comparable with traditional payment-network volumes.
The paper also references regulatory developments including the U.S. GENIUS Act, the EU’s MiCA framework, Hong Kong’s stablecoin regime and Singapore’s regulatory framework.
If machine-to-machine payments expand, stablecoins could be one option for settling those transactions. The research does not indicate how quickly that market could develop.
Tokenized assets could give AI systems financial access
BlackRock also looks at the potential role of tokenized financial assets.
Tokenization creates digital representations of assets such as funds and securities that can be transferred through programmable infrastructure.
For AI systems, standardized digital assets could potentially make it easier to interact with financial products according to predefined rules.
However, tokenization does not remove existing financial requirements. Access to specific assets would still depend on rules covering identity, eligibility, KYC and AML compliance.
Under this model, blockchain infrastructure could handle transfers and settlement, while separate compliance systems determine which participants can access particular products.
Computing capacity could become a financial market
The third area covered by the research is computing capacity.
AI development requires large amounts of computing power for both training and inference. BlackRock examines whether access to that capacity could eventually be priced, financed or represented through digital assets.
The firm cites estimates that revenue from major cloud businesses could reach approximately $1.1 trillion by 2030, representing a 29% compound annual growth rate from 2025 levels.
One potential model would involve digital contracts representing access to a defined amount of computing capacity. Such claims could potentially be transferred, financed or used as collateral.
BlackRock also points to GPU-backed financing as an early example of financial activity developing around AI infrastructure.
AI could select and pay for computing resources
The research further considers whether AI agents could select computing resources based on factors such as price, performance, latency, location and hardware availability.
An automated system could identify its computing requirements, compare available providers and settle the resulting payment.
BlackRock references technologies such as Model Context Protocol (MCP) and Agent2Agent (A2A) for communication between AI systems, alongside payment protocols such as x402.
These technologies are still developing, and their eventual role in blockchain-based financial activity remains uncertain.
BlackRock’s digital asset push extends beyond AI research
The latest research comes alongside BlackRock’s broader expansion into digital-asset products.
Earlier in August, BlackRock expanded its Canadian ETF lineup with the IBQT ETF, which gives Canadian investors diversified equity exposure alongside Bitcoin through a single portfolio.
That product is separate from the AI-focused research, but it adds to BlackRock’s recent activity around integrating digital assets with traditional investment products.
The company’s digital-asset strategy has therefore extended across several areas, including exchange-traded products, tokenization and research into blockchain infrastructure.
Several use cases remain unproven
The research outlines possible connections between AI and digital assets, but each would require further infrastructure and adoption before becoming a significant market.
AI payments would need systems for permissions, identity, compliance and handling failed or disputed transactions.
Computing markets would need standardized ways to measure resources that differ by hardware, location, availability and performance.
Stablecoins and tokenized assets face their own regulatory and interoperability requirements.
BlackRock’s research therefore presents AI-driven payments, tokenized assets and digital compute markets as areas to monitor rather than established sources of blockchain demand.
Whether these applications generate meaningful onchain activity will depend on how quickly the underlying technologies develop and whether businesses adopt them at scale.
Also Read: Grayscale Renames Bitcoin Miners ETF as AI Compute Fund
