AI agents can already research, write code, and run entire workflows. The next step is money. Agents that hold wallets, sign transactions, and spend within human-defined guardrails, with no click-to-approve step in between, are starting to move from demo to production.
That shift forces questions every payments builder now has to answer. Why do stablecoins (programmable, instantly settled, always on) keep coming up as the natural foundation for software that transacts? What does an agent actually need before it touches a cent? And when an agent buys the wrong thing, who is liable?
In this episode of Utila’s Stablecoin Builder Series, Head of Payments Shahar Friedman puts those questions to three teams building the answers: Rain’s Catherine Peng, Tempo’s Brendan Ryan, and Polygon Labs’ James Lawton.Watch on-demand to learn more about:

Works on Rain’s agentic payments initiatives. Previously spent six years at Visa, most recently on Visa Intelligent Commerce.

Leads Polygon Labs’ agentic payments initiatives as part of its work on the money stack.

Leads engineering for Tempo’s AI initiatives, including the Machine Payments Protocol (MPP), the open standard for machine-to-machine payments co-authored by Stripe and Tempo. Previously spent six years at Stripe.

Utila provides fintechs, banks, and enterprises with infrastructure to build and manage stablecoin and digital asset operations.
This session was produced by Utila as part of the Stablecoin Builder Series, a monthly live discussion for founders and operators building in stablecoins.