Internet Money Protocol

Allaire's founding thesis

The thesis is that money, specifically dollar-denominated value, should operate as a first-party protocol of the internet, the way TCP/IP governs how data packets route, SMTP governs how email transmits, or HTTP governs how web pages load. Under this vision, dollars become a data type: programmable, transmissible at near-zero cost, accessible to any developer, and moved without institutional intermediaries.

When Jeremy Allaire founded Circle in 2013, no stablecoin infrastructure existed yet. His founding argument was a structural prediction rather than a product description: "Just as there are protocols for how information moves around, like web protocols or email protocols or voice protocols, there are going to be these new protocols for money on the internet. That will allow very safe forms of dollars to be operated that way. And that was the idea."¹ The analogy is precise. TCP/IP governs how data packets are addressed and routed, with no institution deciding whether a given packet moves. SMTP governs how email is formatted and transmitted, with no company to call in order to send a message. USDC, in this vision, would govern how dollar value is stored and transmitted, with no bank to call in order to move money. The vision was not achievable in 2013, since the technology, programmable blockchain infrastructure beyond Bitcoin, did not yet exist; Allaire treated Bitcoin as layer zero, proof that digital money without a custodian was possible, and bet that the underlying technology would mature within a few years to support a dollar implementation.

First-party data

By 2026, Allaire's language had sharpened to describe what the protocol vision looks like in practice: "The idea behind this is this is general purpose, general architecture money. Make it almost like a first-party form of data on the internet." First-party data is data native to the internet as a medium rather than data hosted by an intermediary and accessed only through that intermediary's own systems. A webpage is first-party data on the internet, readable without going through any single company. A bank wire is not: it requires the bank's own infrastructure at every step. A USDC balance on a blockchain, by contrast, can be read and transacted by any software system with internet access, without the holder's bank, without Circle's per-transaction permission, and without any institutional intermediary at the transaction layer.

Protocol, not product

The protocol framing is strategically different from describing Circle as a company that makes a product called USDC. A product company is valued on unit economics, cost to produce multiplied by price and volume. A protocol company is valued on network effects, where the value of the network to each participant grows with the number of participants using it. Allaire's own description is explicit: Circle runs a stablecoin network, not a stablecoin issuer. That network includes the USDC token itself, the application programming interfaces and software development kits any developer can integrate without calling Circle directly, liquidity spread across exchanges, banks, and decentralized finance protocols globally, and a regulatory compliance layer that makes the network usable by regulated institutions. The protocol analogy also explains a specific competitive dynamic: a bank cannot simply hire away Circle's engineers and build a competing network, in the same way nobody can hire the original authors of SMTP and rebuild email from scratch, because by the time a competing protocol exists the incumbent already holds every integration, all the liquidity, and the accumulated trust.

AI agents as the leading use case

The leading edge of the internet money protocol vision is machine-to-machine micropayments. As Allaire put it, USDC can already be used to make a micropayment between two AI agents in a fraction of a second, at a fraction of a cent, and it is being used that way today. This is a use case existing payment infrastructure cannot serve at all: credit cards carry a two to three percent fee, settle over one to three days, and require human authorization; ACH transfers take one to three days and require bank accounts; wire transfers take hours to days and carry high minimum effective costs. For AI agents to transact with each other, paying for application programming interface calls, computation, data, or services, they need an instrument that settles in milliseconds at fractional-cent amounts, and USDC is currently the only viable one. This makes the rise of an agent-driven economy the most significant structural tailwind for the internet money protocol thesis.

Comparison to other internet protocols

| Protocol | What it governs | Internet-native | Incumbent it replaces | |---|---|---|---| | TCP/IP | Data routing | Yes | Circuit-switched telephony | | HTTP | Web content access | Yes | Print and broadcast media | | SMTP | Email delivery | Yes | Postal mail and fax | | USDC | Dollar value transfer | Becoming | Wire transfer and correspondent banking |

The pattern across all of them is that internet protocols make incumbent institutional intermediaries optional for their respective domain. USDC follows the same pattern for money that HTTP followed for information distribution.

An unusual claim: enlightenment ideals in code

Allaire makes a further, more unusual claim: that the protocol properties of blockchains, permissionless, decentralized, open-source, are not neutral technology choices but expressions of liberal enlightenment ideals such as free speech, open markets, and individual agency, and that this is part of why centrally controlled political systems have struggled to build an equivalent. This is a strong claim that could be tested against the messier history of internet governance itself, where the open internet has been partially centralized or fragmented over time by various sovereign efforts. As a first-order statement, though, it does help explain both the structural resilience of USDC, usable by any developer worldwide, and the difficulty state-controlled entities face in reproducing Circle's competitive position.