The core thesis for crypto in an AI world is that agents need a financial system that doesn't rely on human biometrics for authentication. By equipping agents with their own crypto wallets, platforms like MoonPay are building the infrastructure for them to transact autonomously, making crypto the native currency for a new machine economy.

CZ predicts millions of AI agents will soon transact on our behalf, booking hotels and making micropayments. Traditional banking systems cannot handle this volume, speed, or the KYC requirements for non-human entities, making crypto the only viable payment rail for the agent economy.

As AI agents become sophisticated, they'll need to pay for services. Traditional banking is too slow and fragmented for them. Crypto, as the internet's native money, provides the instant, global, low-fee rails for AI agents to transact with each other and with web services, creating a major new use case.

An emerging narrative suggests that as AI agents increasingly perform autonomous work and transactions 24/7, they will require 'frictionless money' to operate. U.S. dollar-pegged stablecoins are positioned to fill this role, creating a nascent but potentially massive new channel of non-human demand for the currency as these agents transact.

Dragonfly's managing partner argues that attempts to apply crypto to non-financial domains have largely failed. Crypto's core, enduring value is as programmable money. Its next major growth vector will be serving as the native financial rails for AI agents to transact autonomously with each other.

The true addressable market for crypto is not the 8.5 billion humans, but trillions of AI agents needing rails for microtransactions. This 'agents are coming' narrative implies a demand for crypto that is orders of magnitude larger than the much-hyped 'institutions are coming' thesis.

The financial system is unprepared for the coming wave of AI agents. These agents will perform tasks and require payment, creating trillions of micropayments. Current infrastructure from Stripe, Visa, or Mastercard cannot handle this volume, creating a massive opportunity for new protocols to facilitate the 'agent economy'.

AI agents are turning to crypto not just for efficiency, but out of necessity. The traditional financial system is a dead end for non-human entities, as an AI cannot get a credit card or open a bank account. Crypto provides the permissionless financial rails required for AI agents to operate and self-replicate economically.

Companies like Ramp are developing financial AI agents using a tiered autonomy model akin to self-driving cars (L1-L5). By implementing robust guardrails and payment controls first, they can gradually increase an agent's decision-making power. This allows a progression from simple, supervised tasks to fully unsupervised financial operations, mirroring the evolution from highway assist to full self-driving.