Money for machines
Circle CEO Jeremy Allaire explains why his company built Arc, a new blockchain meant to meet regulatory standards for real money and to serve as infrastructure for an economy increasingly run by AI agents. The conversation covers how USDC's network effects and regulated infrastructure give Circle its defensibility.
- Circle built Arc because existing blockchains weren't designed to meet the standards regulators and large financial institutions expect for real money, and Arc's validators must meet high operational, security, and compliance requirements.
- Allaire argues the main users of this infrastructure will likely be AI agents, which shapes design choices around the coming agentic economy.
- USDC's defensibility rests on three things: utility network effects from developers choosing networks where others already build, liquidity network effects across major exchanges, and regulated banking infrastructure in many countries. Circle's Cross-Chain Transfer Protocol is the most widely used interoperability protocol, since USDC is the most transferred digital currency across chains.
Circle, the issuer of USDC, is building Arc, its own blockchain, and the reasoning starts with a decade-old goal: put dollars on the internet and make them work everywhere, something like an HTTP for money. Jeremy Allaire, Circle's cofounder and CEO, says the company found that existing chains weren't built for what regulated financial institutions need. Those institutions want standards around operations, information security, and compliance before they'll put real money and real investment contracts on a network. Arc is designed to meet those standards, with validators who sign service-level agreements and deterministic settlement finality measured in a fraction of a second. Allaire also says the network's main users will likely be AI agents rather than people, which shapes the design toward what he calls an agentic economy. Chris Dixon of a16z crypto, which invested, argues that money has the most obvious network effect of any application, and that dollars' global popularity gives a dollar-denominated chain a real shot at becoming a global money layer.
The more distinctive part of the pitch is governance. Allaire wants the infrastructure run by a geographically diverse set of stakeholders, with major geoeconomic blocs holding real economic stakes and local champions using the network. He says that neutrality is a precondition for internet-scale adoption, since no single country or company should be able to control a system meant to settle economic activity for the whole world. Arc's initial stakeholders include large players in capital markets, asset management, payments, and global banks covering US, European, and Asian markets. Allaire is clear that this is only the start, and that the goal is much broader participation.
On defensibility, the conversation points to Circle's existing position. Allaire notes that the most widely used interoperability protocol across blockchains and assets is Circle's, which gives the company an installed base and a reason for other networks to plug into its standards rather than build their own. The interview never gets into the detailed mechanics of that moat, and the claims about AI-driven demand and future market size are forward-looking. Still, the core argument is concrete: Circle is betting that regulated, multi-jurisdictional, and increasingly machine-driven payments need infrastructure designed for them from the start.