Wall Street's AI Bet Forgot About the Micropayments 🤖

Franklin Templeton, the investment manager overseeing nearly $1.8 trillion in assets, said agentic AI will become the "killer" use case for blockchain adoption, arguing that autonomous AI agents will rely on crypto rails rather than traditional payment networks to transact. Sandy Kaul, the firm's head of digital assets and innovation, published a more-than-1,800-word post on Wednesday making the case that blockchain protocols are better suited than legacy financial infrastructure for machine-to-machine micropayments. "Blockchain will be pivotal in allowing agentic AI to realize its potential for consumer transactions, and the growth of agentic AI is likely to become the 'killer' use case that drives blockchain adoption," Franklin Templeton said.

Kaul's thesis focuses on the difference between generative AI, which produces text, images or code in response to prompts, and agentic AI, which can shop, book travel, buy computing power, compare prices, retrieve data or manage software workflows on a user's behalf without constant supervision. A Capgemini report cited in the paper described the shift as moving AI "from a reactive, conversational chatbot to an autonomous system that can perceive its environment, devise a plan, and execute multi-step tasks to achieve high-level goals without constant human supervision." According to a Bain & Company forecast also cited by Kaul, "AI agents are expected to account for 15% to 25% of all U.S. e-commerce sales by 2030."

Kaul said traditional payment networks are too slow and expensive for agentic commerce. While bitcoin processes roughly 7 transactions per second and Ethereum roughly 75 TPS, the post noted that newer high-speed chains record maximum speeds of 12,933 TPS on the Aptos chain, 6,284 TPS on Solana and 3,252 TPS on the BNB Chain. By comparison, the Visa network processes 1,700 to 10,000 transactions per second in normal operations but only records transactions in that window, with settlement taking one to three business days. Kaul said networks such as Aptos, Solana and BNB Chain are more suited for the agentic economy because they settle transactions in seconds. Circle CEO Jeremy Allaire has separately argued that AI agents and blockchain are converging into a single economic system where software can transact, coordinate and exchange value autonomously.

The infrastructure is already being built. A joint report published last Wednesday by Visa and investment thesis platform Artemis argued that traditional cards built for low-frequency human commerce are insufficient for AI agents, which need infrastructure with near-zero fees and faster settlement to make agentic micropayments commercially viable. Visa's crypto division and Stripe-backed Tempo both launched AI tools in March, with Visa's allowing AI agents to make same-day payments. Coinbase's x402 payment protocol has processed $15 million in adjusted volume across more than 109 million adjusted transactions since launching in May 2025, according to the Visa-Artemis report. Robinhood launched AI-powered investing tools in May that let agents trade stocks and make purchases for users, and CEO Vlad Tenev has said AI agents will eventually rival the capabilities of human traders.

Industry groups are forming around the same thesis. The x402 Foundation, which includes 40 organizations such as Visa, Mastercard and AWS building open payment rails for AI, launched formally on July 14. Google unveiled a payment protocol for agents in 2025 backed by the Ethereum Foundation. OpenAI and Anthropic are racing to build increasingly autonomous systems that can navigate software and complete complex tasks on their own, a trajectory Kaul said investors ignore at their peril if they continue to frame AI exposure solely as a chipmaker and cloud trade.