Franklin Templeton: Blockchain Is the Next AI Trade as Autonomous Agents Start Spending

Franklin Templeton's head of digital assets and innovation argues that AI portfolios focused on Nvidia and cloud infrastructure are overlooking the payment rails that autonomous agents will require to transact with one another.

The Thesis: AI Portfolios Are Missing Blockchain

Franklin Templeton's head of digital assets and innovation, Sandy Kaul, published an argument on July 22, 2026, as reported by CoinDesk, that institutional AI portfolios may have already missed the semiconductor stage and should now consider blockchain networks and crypto assets. Kaul's argument centers on agentic AI — autonomous systems built to complete tasks with little human input — and the payment infrastructure those systems will require.

According to CoinDesk's reporting, Kaul argued that many transactions between AI agents could be worth only fractions of a cent, such as paying for an API call, a second of computing power, or access to a dataset. Traditional payment networks, Kaul argued, become expensive when fees cost more than the transaction itself.

Why Blockchain, Not Banks

According to the CoinDesk report, Kaul argued that public blockchain networks are better suited to machine-to-machine payments because they offer programmable transactions, cryptographic identity, and near-instant settlement. Instead of relying on banks or card networks, AI agents could hold digital assets and pay one another directly over blockchain rails.

The investment implication Kaul drew, as reported by CoinDesk: if agentic AI transactions happen at scale, demand for blockchain networks could grow alongside AI adoption. Since agents would need native cryptocurrencies to pay network fees, Kaul argued that rising transaction volumes could increase demand for those tokens while generating more revenue for developer incentives, network security, and decentralized applications.

Circle CEO Makes a Parallel Case

The CoinDesk report also covers a parallel argument from Circle CEO Jeremy Allaire, who Kaul's thesis aligns with. According to CoinDesk's reporting of a paper Allaire published, he argued that AI is driving the cost of knowledge work toward zero while blockchain and programmable digital money are doing the same for payments, settlement, and coordination.

As businesses rely more heavily on specialized AI agents, CoinDesk reported Allaire's view that those agents will become economic actors that buy services, hire other agents, and exchange value autonomously. Blockchain networks, digital identities, and programmable money would provide the infrastructure needed to support those interactions at internet scale. Allaire's argument also extended to pricing models — software pricing could shift from monthly subscriptions to pay-per-task models as AI agents become both buyers and sellers of digital services.

What This Means for Builders

The Kaul-Allaire framework describes a transition from AI as a productivity tool to AI as an economic actor. For infrastructure builders and tokenization platforms, the implication is that the demand for digital payment rails may not be primarily driven by human retail users. It may be driven by autonomous agents executing millions of small transactions daily.

Kaul's argument, as reported by CoinDesk, is not that blockchain replaces AI infrastructure — it is that blockchain is the settlement layer for AI's next phase of economic activity. The two build on each other rather than competing for investment capital.

Sources

• CoinDesk — "Forget Nvidia: The next big AI trade could be crypto and blockchain" — July 22, 2026: coindesk.com