Franklin Templeton says agentic AI might turn into the “killer use case” that blockchain and crypto have been trying to find over a few years, as autonomous AI brokers start to require on the spot, programmable, and verifiable cost infrastructure with out steady human intervention.
This view was introduced by Sandy Kaul, Head of Digital Belongings and Innovation at Franklin Templeton, in an evaluation printed in July 2026, amid cost corporations, AI corporations, and crypto infrastructure suppliers constructing transaction layers for machines. Stripe, the Linux Basis, and OpenAI have all taken steps associated to funds for AI brokers, indicating that agentic commerce is shifting from a technological idea to its preliminary infrastructure-shaping stage.
— Franklin Templeton Digital Belongings (@FTDA_US) July 21, 2026
Franklin Templeton’s Thesis
Sandy Kaul, Head of Digital Belongings and Innovation at Franklin Templeton, doesn’t view agentic AI merely as a brand new utility layer of synthetic intelligence, however as a check for whether or not blockchain can serve actual financial wants past digital asset buying and selling.
In line with Kaul, that is the place blockchain might discover a extra sensible function in comparison with earlier speculative cycles. An AI agent could have to pay for information, name APIs, buy software program, e-book companies, or execute a number of small transactions in a process chain. Such transactions could also be too small, too quick, or too frequent to move by conventional cost processes designed for human customers.
The asset supervisor has constructed a clearer presence within the tokenization area by the Franklin OnChain U.S. Authorities Cash Fund, linked to the BENJI ecosystem. The fund recorded whole internet property of $753.24 million as of June 30, 2026, exhibiting that blockchain is being utilized in a standard monetary product fairly than remaining solely in crypto experiments.
The Case for Blockchain-Primarily based AI Funds
When AI brokers execute transactions autonomously, the cost system should know whom the agent represents, how a lot it’s licensed to spend, and depart an auditable transaction historical past when wanted.
This hole makes blockchain and stablecoins a notable choice for funds between AI brokers. The expertise can document transactions in a clear and programmable method, whereas stablecoins present a settlement unit much less unstable than typical crypto property. For small quantities, resembling an agent paying a price for a single API name or information entry occasion, a pay-per-use on the spot cost mannequin could also be extra appropriate than conventional subscriptions or invoices.
Crucially, blockchain doesn’t essentially have to interchange Visa, Mastercard, or banking methods. A extra sensible use case is serving as a supplementary cost layer for transactions that present infrastructure processes sub-optimally: machine-to-machine funds, micropayments, pay-per-use APIs, cross-border settlements, and conditional automated transactions.
Market Indicators Behind the Thesis
Franklin Templeton provided this evaluation at a time when the market had already proven extra concrete alerts relating to funds for AI brokers. In March 2026, Stripe launched the Machine Funds Protocol, an open normal co-developed by Stripe and Tempo for brokers to pay for sources, APIs, or companies by way of HTTP endpoints whereas connecting to Stripe’s present cost infrastructure.
By July 2026, the Linux Basis introduced that the x402 Basis formally went stay after Coinbase contributed the x402 protocol. The muse has 40 members, together with Coinbase, Stripe, Visa, Mastercard, Google, AWS, and Shopify, meaning to standardize internet-native funds for AI brokers, APIs, and functions.
OpenAI can be bringing agentic commerce nearer to mainstream customers. Instantaneous Checkout in ChatGPT, constructed with Stripe on the Agentic Commerce Protocol, permits U.S. customers to buy immediately from Etsy sellers inside the chat, with over a million Shopify retailers introduced to be supported later. With over 700 million weekly ChatGPT customers, conversational checkout might turn into a industrial channel price watching.
Market forecasts are additionally reinforcing this story. McKinsey estimates that AI brokers might orchestrate $3 trillion to $5 trillion in world client transactions by 2030, counting bodily items alone. Gartner predicts that by 2028, 33% of enterprise software program functions will incorporate agentic AI, and no less than 15% of day-to-day work choices may very well be made autonomously by agentic AI.
Dangers and What Comes Subsequent
Nonetheless, “killer use case” stays a thesis that must be verified by real-world adoption. As AI brokers start executing transactions autonomously, the troublesome query isn’t just whether or not the cost expertise works, however who bears legal responsibility if an agent makes a mistaken buy, will get scammed, or exceeds its allotted limits.
These dangers will immediately have an effect on the tempo of deployment. Gartner has warned that greater than 40% of agentic AI tasks may very well be canceled earlier than the top of 2027 attributable to rising prices, unclear enterprise worth, or inadequate danger controls.
Within the quick time period, the important thing facet to observe is whether or not MPP, x402, and checkout fashions in AI apps enter enterprise workflows and client commerce. If AI brokers generate actual transaction quantity, notably in small, automated funds, blockchain may have a clearer foundation to be seen as infrastructure for a brand new layer of commerce. If not, this “killer use case” will stay a lovely concept fairly than a confirmed adoption story.
