Sandy Kaul, head of digital assets and innovation at Franklin Templeton, argued Wednesday that blockchain protocols are better suited than legacy payment networks for machine-to-machine micropayments in an autonomous AI agent economy.
Kaul posted a 1,800-word analysis on X questioning whether traditional investment strategies will work as AI agents proliferate. “To capture the AI growth opportunity today, most investors buy shares of AI-aligned companies and related verticals. But will the same playbook work for agentic AI,” Kaul wrote.
The core problem, according to Kaul’s argument, centers on settlement speed and fees. Legacy card networks like Visa settle transactions in 1-3 business days and carry high fees unsuitable for frequent, low-value transfers between machines. Blockchain networks, by contrast, settle transactions in seconds, making them more efficient for AI agents that need to execute thousands of micropayments daily.
Visa and Artemis, an investment thesis platform, published a joint report last Wednesday supporting this view. The companies argued that traditional cards, built for low-frequency human commerce, are structurally insufficient for AI agents. Visa’s crypto division and Tempo, a payments tool backed by Stripe, launched AI tools in March 2026 that allow AI agents to make same-day payments.
Coinbase’s x402 payment protocol, launched in May 2025, has processed 109 million adjusted transactions and $15 million in adjusted volume since launch, demonstrating early traction in blockchain-based machine-to-machine settlement.
Kaul’s thesis does not specify which blockchain networks, such as Aptos, Solana, or BNB Chain, are best positioned for agentic payments, nor does it detail specific use cases beyond payment settlement infrastructure.
