Franklin Templeton: Blockchain Essential for Agentic AI Commerce

By Khal · · 2 min read

Editorial illustration for: Franklin Templeton argues blockchain better suited for agentic AI than traditional rails
Image generated for editorial use · Generated via fal.ai FLUX — editorial license · Published Jul 22, 2026, 7:53 p.m. (1 hour ago)

In brief

  • Franklin Templeton's Sandy Kaul published a paper arguing agentic AI requires blockchain infrastructure to function at scale
  • AI agents projected to handle 15-25% of U.S. e-commerce by 2030, exceeding traditional payment network capacity
  • Blockchain networks like Solana and Aptos process thousands of transactions per second versus Visa's 1,700-10,000 TPS
  • Kaul argues investors must hold cryptocurrencies to capture value from decentralized agentic commerce networks

Why the infrastructure matters

Agentic AI agents can shop, book, and pay for things autonomously, and the market is nascent. Per a Bain & Company forecast cited by Kaul, AI agents are expected to account for 15% to 25% of all U.S. e-commerce sales by 2030. That volume requires infrastructure built differently than what exists today.

Traditional payment networks—built for human-scale, human-speed commerce—can't handle the transaction volume required for widespread agentic AI adoption. The math is stark. Bitcoin processes approximately 7 transactions per second and Ethereum processes approximately 75 transactions per second. The Visa network processes 1,700 to 10,000 transactions per second in normal operations, but settlement takes 1-3 business days, whereas blockchains record and settle transactions simultaneously.

Kaul's argument hinges on speed and finality. Visa's settlement delay reflects a business choice, not a technical constraint. Layer-2 solutions and payment channels already enable sub-second settlement on Ethereum at lower cost than base-layer transactions. Cost-per-transaction, not speed alone, may ultimately determine whether blockchain or traditional infrastructure wins for agentic AI payments.

The infrastructure is already moving

Newer high-speed chains are recording maximum speeds ranging from 12,933 transactions per second (TPS) on the Aptos chain, 6,284 TPS on Solana, and 3,252 TPS on BNB Chain. Meanwhile, major tech companies are preparing their own infrastructure. Coinbase already launched tools that let AI agents trade and pay autonomously, and Google unveiled a payment protocol for agents in 2025, backed by the Ethereum Foundation.

But the narrative isn't simply "blockchain wins." The x402 Foundation—40 organizations including Visa, Mastercard, and AWS—launched formally to build open payment rails for AI. The x402 protocol revives a forgotten HTTP status code from 1991, originally reserved for web payments, and puts it to work letting software pay software directly over the internet. This is a competing infrastructure play built on traditional web standards, not blockchain. Kaul's argument is one perspective in an ongoing infrastructure debate, not an inevitable outcome.

What investors should know

Agentic commerce—AI systems transacting autonomously on behalf of humans, from buying cloud compute to booking flights—is projected at $3 trillion to $5 trillion by 2030 according to a McKinsey & Company report. The scale is real.

Kaul's thesis is that investors will need to buy the cryptocurrencies and alt coins being issued by decentralized networks and businesses to capture the value of those entities. This is his stated institutional perspective, not an endorsement from LeoDex. Kaul's argument is speculative and reflects one perspective on how infrastructure will evolve; readers should conduct independent due diligence before making investment decisions.

Related stories