TL;DR:
Anchorage Digital, the federally chartered crypto bank, is betting that artificial intelligence agents will become full participants in the global financial system.
Nathan McCauley, co-founder and CEO of the firm, presented this view at the Wyoming Blockchain Symposium 2026, where he described these autonomous systems as future “first-class economic actors” capable of operating interchangeably across cash, card, and cryptocurrency rails.
McCauley noted that the evolution of AI agents is not limited to advisory functions, but scales toward the direct execution of transactions. “They’re not just going to advise, they’re going to carry out the transactions,” he stated.
He also added that this requires a set of controls and structures that allow them to interact both on-chain and with cards and traditional rails. As a cultural reference, the executive evoked figures from science fiction — from the domestic robots of The Jetsons to Jarvis, Iron Man’s assistant — to illustrate the type of autonomous bots he envisions.
In May, Anchorage launched its Agentic Banking platform, an institutional environment designed to allow organizations to fund and control AI systems that operate on their behalf. The infrastructure features a trust, governance, and settlement layer, along with a proprietary identification system called “know-your-agent“, adapted to the compliance requirements involved in delegating financial decisions to autonomous systems.
McCauley was explicit that the concept of “agentic payments” falls short. According to the executive, agents need full bank accounts: the ability to receive funds, manage them, and spend them in a channel-agnostic manner. This distinction represents a crucial conceptual difference from the point-to-point payment solutions that currently dominate discussions around AI and finance.
The CEO also anticipated a significant increase in microtransactions driven by autonomous bots, and noted that blockchain networks are particularly well suited for that scenario. He compared gas fees to a primitive form of micropayment for service, and projected that bots could pay for individual API calls rather than subscribing to full services, creating a new high-frequency, low-value transactional commerce model.
