TLDR: BlackRock says blockchains are moving from speculative trading tools to AI agent settlement rails. Stablecoin transaction volume topped $11 trillion in 2025, rivaling major card network
BlackRock has released a new whitepaper stating that blockchains will evolve from speculative rails into decentralized value settlement networks for AI agents.
The global asset manager argues that autonomous AI systems require machine-native financial infrastructure to operate independently.
According to BlackRock, this shift marks a fundamental change in how blockchain networks function within the broader financial system going forward.
BlackRock states that blockchain technology has historically served as a vehicle for human speculation and trading activity.
The report argues this role is changing as AI agents gain the ability to interact with financial networks directly. Agentic AI systems can now plan and execute multistep tasks with limited human involvement.
The paper explains that blockchains provide programmable infrastructure connecting machine intelligence with economic activity.
This infrastructure allows AI systems to move beyond content generation into real-world financial action. These actions include making purchases and initiating transactions without manual approval at each step.
BlackRock draws a parallel between AI tokenization and blockchain tokenization to support this argument. Language models convert text into numerical tokens for processing and interpretation.
Blockchains apply a similar process, converting assets into standardized digital tokens for machine-verifiable transfer and settlement.
This shared architecture, according to BlackRock, gives AI agents a more direct interface with blockchain data. Traditional financial systems remain fragmented and often require bespoke integrations for each connection. Blockchain-based settlement networks reduce this complexity as machine-to-machine transactions scale upward.
BlackRock identifies several protocols building the foundation for blockchain-based agent settlement. These include x402, developed by Coinbase, which uses the HTTP 402 status code for automated payments. The protocol remains blockchain agnostic but relies heavily on stablecoins for early transactions.
The report also names the Machine Payments Protocol from Stripe and Tempo as another settlement tool. This protocol enables payments for APIs and other resources with flexible stablecoin settlement options. Stripe and OpenAI jointly developed the Agentic Commerce Protocol for programmatic checkout processes.
Samara Cohen, BlackRock’s Global Head of Market Development, has separately described stablecoins as “becoming the bridge between traditional finance and digital liquidity.” That framing aligns closely with the settlement network argument made throughout the AI agent whitepaper.
Elsewhere in the industry, Sandy Kaul of Franklin Templeton posted on X questioning whether buying AI-aligned equities alone would work for capturing “the same playbook” in the agentic AI economy. His comments echo BlackRock’s broader thesis that blockchain-native exposure may matter as much as equity exposure.
BlackRock points to stablecoin transaction data as early evidence supporting the settlement network thesis. Adjusted stablecoin transaction volume exceeded $11 trillion in 2025 according to the report. This figure places stablecoin activity within a similar range as Visa and Mastercard combined.
The report states that stablecoins now represent the largest category of tokenized real-world assets. Circulating stablecoin market capitalization surpassed $300 billion as of September 2026. Price stability makes these tokens useful for predictable settlement across agent-driven transactions and workflows.
Stablecoin volume grew at an 80% compound annual rate between 2020 and 2025. ACH volume grew roughly 8.5% over the same period by comparison. BlackRock frames this gap as a signal of accelerating machine-native settlement demand.
Regulatory clarity is also cited as a factor supporting this transition toward settlement networks. The GENIUS Act in the United States and MiCA in Europe provide clearer frameworks. Hong Kong and Singapore have introduced their own stablecoin licensing regimes as well.
BlackRock extends its settlement network argument to compute, the processing power behind AI systems. As agents become more capable, standardized claims on compute capacity could become a digital asset. These claims could be transferred, financed, and settled through programmable blockchain infrastructure.
The report estimates that AI capital spending could exceed $5 trillion between 2025 and 2030. Combined cloud revenue from major hyperscalers could reach $1.1 trillion by 2030. This growth increases the need for pricing, hedging, and settlement tools around compute.
BlackRock suggests that on-chain tokenized markets could enable region-specific and hardware-specific compute contracts.
Agents could query real-time pricing data and provision resources through decentralized settlement rails. This process would rely on x402 for per-use or per-job payment settlement.
Stripe’s planned acquisition of OpenRouter is cited as an early signal of this convergence. BlackRock views this deal as evidence that compute procurement and blockchain settlement are merging into a single financial layer supporting AI agents.
The post BlackRock: AI Agents Will Turn Blockchains into Machine-Native Settlement Rails appeared first on Blockonomi.
