Blockchain Networks Move Into Focus as AI Agents Create a New Crypto Investment Thesis

What to Know

• Franklin Templeton’s Sandy Kaul says investors looking beyond AI chipmakers and cloud companies should consider blockchain networks and crypto assets.
• The thesis centers on agentic AI, where autonomous software can complete tasks and potentially transact with other agents with little human input.
• Machine-to-machine payments could involve tiny transactions, including API access, computing power, data retrieval and software workflow execution.
• Traditional financial networks may be poorly suited to transactions worth fractions of a cent when fees exceed the value being exchanged.
• Public blockchains may offer programmable transactions, cryptographic identity and near-instant settlement for AI-driven commerce.
• Circle CEO Jeremy Allaire has argued that AI agents and blockchain could converge into a shared economic system based on programmable money and autonomous coordination.
• If AI agents begin using native cryptocurrencies to pay network fees, higher transaction activity could support demand for blockchain tokens.
• The investment debate is shifting from chips and data centers toward the infrastructure that may allow software agents to exchange value directly.

AI Trade Expands Beyond Chips and Cloud

Artificial intelligence has become one of the most powerful investment themes in global markets, with much of the attention focused on semiconductor firms, hyperscale cloud providers and data center infrastructure. That focus is understandable because advanced AI models require enormous computing resources, specialized hardware and scalable cloud environments. However, a growing part of the market conversation is now moving toward a different question: what happens when AI systems stop being only tools for generating content and begin acting as economic participants?

Franklin Templeton’s Sandy Kaul has placed blockchain networks and crypto assets at the center of that discussion. Her argument is not that blockchain replaces the existing AI infrastructure boom, but that it may become part of the next layer of the trade. As autonomous AI agents become more capable, they may need to pay for services, acquire information, reserve computing resources, compare prices, trigger workflows and exchange value with other software systems. That activity could require financial rails that are faster, cheaper and more programmable than many legacy payment systems.

For investors, the idea is important because the first stage of the AI trade has been relatively concentrated. Public technology companies tied to model training, cloud deployment and hardware supply chains have dominated institutional flows. Kaul’s view suggests that the next stage may be less about the machines that train AI models and more about the networks that allow AI agents to operate economically once those models become autonomous.

Why Agentic AI Changes the Payment Debate

Agentic AI differs from generative AI in a critical way. Generative AI responds to prompts by producing text, images, code or other outputs. Agentic AI is designed to complete tasks with far less human direction. An AI agent could book travel, compare vendors, retrieve data, manage software processes, purchase computing capacity or coordinate with other agents to execute a broader objective.

That distinction matters for payments. A chatbot that answers a question does not necessarily need to move money. An autonomous agent acting on behalf of a person, business or application may need to make many small payments while completing a task. Those payments could cover an API call, access to a dataset, a short burst of computing power or the use of another specialized software agent.

Many of those transactions could be worth only fractions of a cent. Traditional financial infrastructure is not typically optimized for that type of machine-speed, ultra-low-value activity. Card networks, bank transfers and other conventional rails can involve costs, settlement delays or operational complexity that make very small transactions uneconomic. If the fee is larger than the value being transferred, the payment model breaks down.

This is where blockchain networks enter the investment thesis. Public blockchains can support programmable transactions, digital asset settlement and cryptographic identity in a way that may be more compatible with autonomous software. In this framework, an AI agent could hold digital assets, prove its identity through cryptographic credentials and pay another agent directly for a service without relying on a traditional intermediary at every step.

Blockchain as Machine-to-Machine Payment Infrastructure

Kaul’s argument centers on blockchains as low-cost, programmable rails for machine-to-machine commerce. The key issue is not simply that crypto can move value. It is that blockchain networks can combine value transfer with software logic. Payments can be conditional, automated and embedded directly into digital workflows. That makes them potentially useful for a world where AI agents are initiating and completing transactions without constant human approval.

Near-instant settlement is another important part of the thesis. If agents are coordinating in real time, waiting for delayed settlement can create friction. Blockchain-based systems can allow transactions to finalize quickly, while also creating transparent records of activity. Cryptographic identity may also help distinguish legitimate agents, applications and services in an automated environment where trust becomes harder to establish through traditional human relationships.

If this activity grows meaningfully, network usage could rise alongside AI adoption. Many public blockchains require native cryptocurrencies to pay transaction fees. Under Kaul’s framework, higher transaction volumes from AI agents could increase demand for those tokens. At the same time, fees generated by network activity could support developer incentives, security budgets and decentralized applications built around machine-driven commerce.

This does not mean every blockchain asset would benefit equally. Investors would still need to evaluate scalability, transaction costs, developer adoption, security, liquidity and real-world usage. The broader point is that AI-driven economic activity may create a new demand source for blockchain infrastructure, especially if autonomous agents require settlement systems that can handle high volumes of small, programmable payments.

Circle’s View: AI and Blockchain as One Economic Shift

Circle CEO Jeremy Allaire has made a related argument, framing the convergence of agentic AI and blockchain as one technological shift rather than two separate trends. In that view, AI reduces the cost of knowledge work, while blockchain and programmable digital money reduce the cost of payments, settlement and coordination. Together, they could create an internet-native economy where software agents not only generate outputs but also transact, negotiate and coordinate value exchange.

Allaire’s vision extends beyond payments. As businesses rely more heavily on specialized AI agents, those agents could become economic actors that buy services, hire other agents and manage workflows. Blockchain networks, digital identities and programmable money could provide the infrastructure for those interactions at scale. In such a system, software may increasingly operate as both a buyer and seller of digital services.

That has implications for business models. Software pricing could shift away from conventional monthly subscriptions and toward pay-per-task structures, particularly when AI agents consume services on demand. If an agent needs a data query, a code review, a compliance check or a specific computational task, it may pay only for that unit of work. Blockchain-based rails could make those transactions easier to automate and settle.

Allaire has also argued that AI-native companies could increasingly operate on-chain, with tokens representing ownership and governance. That remains a forward-looking view and depends on adoption, regulation and technological execution. Still, it reflects a larger market idea: if AI agents become important participants in the digital economy, the infrastructure for identity, coordination and payment may need to evolve as well.

Early Signs of AI Agents Entering Finance

The shift toward agentic AI is already beginning to appear in financial technology. Robinhood launched AI-powered investing tools in May that allow agents to trade stocks and make purchases for users. CEO Vlad Tenev has said AI agents may eventually rival the capabilities of human traders. OpenAI and Anthropic are also racing to build more autonomous systems that can navigate software and complete complex tasks.

These developments show why the payment question is becoming more urgent. If agents can search, compare, decide and execute, then payments become part of the workflow. A software agent that can choose a service provider but cannot settle payment efficiently remains limited. The more autonomy agents gain, the more important it becomes to define how they hold value, authenticate themselves and interact with other economic actors.

Crypto advocates see this as a natural fit for blockchain networks. Skeptics may argue that traditional payment providers can adapt, or that private systems may capture much of the activity. The outcome is not guaranteed. However, the market debate has clearly expanded. AI is no longer just a story about graphics processors, cloud capacity and model development. It is also becoming a story about how autonomous software may participate in markets.

Investment Implications for Crypto Markets

For crypto investors, the AI-agent thesis introduces a new potential demand narrative. Instead of focusing only on speculative trading, store-of-value arguments or decentralized finance, market participants are increasingly looking at whether blockchains can serve as infrastructure for automated digital commerce. If AI agents begin paying one another for services, data and computing power, the networks that process those payments could see higher usage.

That could matter for native tokens because many blockchains require tokens to pay network fees. Greater demand for blockspace may translate into more fee activity. Those fees can help support validator economics, network security and ecosystem development. Decentralized applications may also benefit if they provide services that agents need, such as data access, identity verification, transaction routing or task execution.

Still, investors should treat the thesis as a developing market narrative rather than an established outcome. Agentic AI is advancing quickly, but widespread autonomous machine-to-machine commerce remains an emerging area. Regulatory questions, security risks, user permissions, identity standards and blockchain scalability all remain important constraints. A future where AI agents transact widely on public networks would require technical reliability and trust frameworks that can operate at scale.

Even with those caveats, the discussion marks a notable shift in how some institutions are thinking about crypto. Blockchain networks are being framed not only as financial speculation, but as potential infrastructure for the next phase of AI adoption. If autonomous agents become a meaningful part of the economy, crypto assets tied to useful, secure and scalable networks may become part of broader AI portfolio construction.

Frequently Asked Questions (FAQs)

Why are blockchain networks being linked to AI investing?

Blockchain networks are being linked to AI investing because autonomous AI agents may need programmable payment rails to buy services, data and computing resources. Some market participants believe public blockchains could support those machine-to-machine transactions more efficiently than traditional payment systems.

What is agentic AI?

Agentic AI refers to artificial intelligence systems designed to complete tasks with limited human input. Unlike generative AI, which responds to prompts by creating content, agentic AI can take steps toward a goal, such as comparing options, booking services, retrieving data or executing workflows.

Why might traditional payment networks struggle with AI agent transactions?

Many AI-agent transactions could be extremely small, potentially worth fractions of a cent. Traditional payment systems may be inefficient for these use cases when fees, settlement delays or operational friction are larger than the value being transferred.

How could crypto assets benefit from AI agents?

If AI agents use public blockchains to transact, they may need native cryptocurrencies to pay network fees. Higher transaction volumes could increase demand for those tokens and generate more revenue for network security, developer incentives and decentralized applications.

Does this mean all cryptocurrencies will benefit from AI adoption?

No. The thesis does not imply that every cryptocurrency will benefit. Networks would still need to prove they can offer low costs, security, scalability, developer activity and real usage for machine-to-machine commerce.

What role does programmable money play in this trend?

Programmable money can allow payments to be automated, conditional and embedded into software workflows. That feature may be useful for AI agents that need to pay for specific tasks, services or data without requiring manual processing for every transaction.

How does Circle’s view fit into the AI and blockchain thesis?

Circle CEO Jeremy Allaire has argued that agentic AI and blockchain are converging into a single economic system. In that view, AI lowers the cost of knowledge work, while blockchain and programmable digital money lower the cost of payments, settlement and coordination.

Is the AI-agent blockchain thesis already proven?

No. The thesis remains forward-looking. AI agents are becoming more capable, but large-scale autonomous commerce on public blockchains still depends on adoption, regulation, security, identity standards and technical performance.

Why should investors watch this trend?

Investors should watch this trend because it may broaden the AI trade beyond chipmakers, cloud providers and data centers. If autonomous agents become economic actors, blockchain networks could become part of the infrastructure that supports the next stage of AI growth.

Photo by DS stories on Pexels