Franklin Templeton says blockchain is the next AI infrastructure play
As autonomous AI agents begin making sub-cent transactions, Franklin Templeton and Circle argue blockchain networks will replace traditional payment rails as the essential infrastructure of the AI economy.
Institutional investors who have poured capital into AI chipmakers and cloud providers may need to look further down the technology stack. Franklin Templeton’s head of digital assets, Sandy Kaul, argues that blockchain networks and crypto assets will power the next phase of artificial intelligence growth.
The thesis hinges on the rise of agentic AI, systems designed to execute tasks like booking travel or buying computing power with minimal human oversight. Robinhood has already introduced AI tools that trade stocks and make purchases, with CEO Vlad Tenev predicting these agents will eventually rival human traders. As these systems proliferate, they will need to transact with one another autonomously.
This machine-to-machine commerce creates a structural problem for traditional finance. Many of these interactions, such as paying for a single API call or a fraction of a second of compute time, will be worth less than a cent. When transaction fees exceed the value of the transfer, conventional banking and card networks become practically unusable.
Kaul argues that public blockchains are better suited for this environment because they offer programmable transactions, cryptographic identity, and near-instant settlement. Market data supports the idea that blockchain rails are already scaling to handle heavy transaction volume. In the second quarter, TRON captured 28.7% stablecoin dominance as its USDT supply hit an all-time high of $89 billion, generating $89 million in protocol fees.
Circle CEO Jeremy Allaire shares this perspective, framing AI and blockchain as a single converging technological shift. In a recent paper, Allaire wrote that AI is driving the cost of knowledge work toward zero, while programmable digital money is doing the same for payments and settlement. He expects AI agents to act as independent economic actors that hire other agents and exchange value.
This dynamic could force a broader change in how software is priced and owned. Allaire suggested that AI-native companies may increasingly operate entirely on-chain, using tokens for ownership and governance. Furthermore, software pricing could shift from monthly subscriptions to pay-per-task models, with AI agents serving as both buyers and sellers.
For portfolio managers, the implication is a potential expansion of what counts as an AI investment. If autonomous agents become a meaningful part of the economy, the native cryptocurrencies required to pay network fees could see rising demand. That would generate revenue for network security and developer incentives, turning blockchain infrastructure into a direct proxy for AI adoption.
