Sandy Kaul, @FTDA_US head of digital assets and innovation at Franklin Templeton, argues that autonomous AI agents represent a structural shift in how economic activity flows, and that legacy
Sandy Kaul, @FTDA_US head of digital assets and innovation at Franklin Templeton, argues that autonomous AI agents represent a structural shift in how economic activity flows, and that legacy payment infrastructure is not built to keep up.
The core problem is one of economics. Standard card networks charge roughly 2% to 3% plus a flat fee per payment, making tiny machine-to-machine transactions commercially unviable. Card networks also settle in one to three business days, a timeline that is simply incompatible with software agents transacting in seconds at near-zero cost. Legacy payment rails with high fees and slow settlement times do not work for micropayments. AI agents also face a more fundamental barrier: they cannot open bank accounts or access financial services that carry strict KYC requirements.
Kaul's argument is that blockchain networks fill that gap directly. Blockchains can settle sub-cent transactions in seconds and automatically record them, making them the natural infrastructure for an agent-driven economy. She singles out @solana, @Aptos, and @BNBCHAIN as networks already suited to that role. Those networks settle transactions in seconds, faster than the one-to-three business-day settlement time of the Visa network.
The activity is already showing up in on-chain data. The x402 protocol, incubated by Coinbase and Cloudflare and now stewarded by the Linux Foundation, has processed roughly $15 million in adjusted volume across 109.6 million transactions since its May 2025 launch.On x402, the average payment is a fraction of a cent, and a fixed card fee on a transaction that small would cost far more than the payment itself.
A joint report from Visa and Artemis, titled "Agentic Payments from the Ground Up," frames the moment as an inflection point. The report found that AI agents are initiating a foundational change in commerce, but current infrastructure gaps are limiting mainstream adoption. The volume figures are still modest by any macro standard, but the transaction frequency tells a different story. Tiny money, enormous frequency.
For investors, Kaul's broader point is a strategic one. Estimates suggest agentic commerce could reach $3 to $5 trillion by 2030, and the playbook of buying shares in AI-aligned companies may not capture that opportunity the same way exposure to the underlying blockchain rails could.
Sources:Franklin Templeton: Agentic AI, The Killer Use Case for Blockchain and CryptoVisa and Artemis: Agentic Payments from the Ground UpCoinTelegraph: Agentic AI is Next Killer Use Case for Blockchain, Franklin Templeton
