THE FORGOTTEN STATUS CODE
HTTP was specified in 1989 with a status code reserved for paid content: 402 Payment Required. It was never implemented. The web's founders expected micropayments would emerge; instead, advertising filled the vacuum and 402 sat unused for thirty-five years.
WHY MICROPAYMENTS FAILED
Every attempt at web micropayments — Millicent (1995), DigiCash, Flooz, Beenz — collapsed on the same wall: card networks charge ~30¢ minimum per transaction, making sub-dollar payments structurally unprofitable. Stablecoins on L2 rails settle for fractions of a cent, removing the floor that killed every prior attempt.
THE GRAND BARGAIN
The modern web runs on an implicit deal: publishers serve content free, users tolerate ads and tracking, advertisers pay publishers via Google and Meta. Roughly $700bn flows through this loop annually. AI agents break it by reading the content and skipping the ads — the consideration the publisher was paid for never gets delivered.
WHY STABLECOINS, NOT CARDS
A stablecoin is a token pegged 1:1 to the dollar, redeemable against reserves of Treasuries and bank deposits. Settlement is final in seconds, fees are denominated in cents on L2s, and there is no chargeback layer — exactly the properties a machine-to-machine payment needs and exactly what card networks cannot offer.
THE AGENT ECONOMICS PROBLEM
An AI agent answering a question may hit fifty pages to compose one reply. At card-network economics that is impossible. At x402 economics — say a tenth of a cent per fetch — it is a rounding error for the agent and a real revenue stream for the publisher. The unit economics flip from impossible to obvious.
THE PRECEDENT THAT WORKED
The one micropayment system that ever scaled was the iTunes Store: 99¢ songs, paid via card, only viable because Apple aggregated thousands of transactions before settling with the network. x402 inverts that — instead of aggregating to fit the rails, it changes the rails so each request can settle on its own.