x402 settlement census — what hosts deliver, round by round https://councilof.ai/ A fixed population of conformant x402 hosts, paid once per round from a wallet we control, with the outcome of every purchase recorded and the rounds diffed. Derived from the committed rows; measurement, not certification; no host is ranked or recommended. en-gb Round 2026-09-06 — 100 delivered, 213 refused of 316 hosts paid https://councilof.ai/interop/x402-census/rounds/2026-09-06.json https://councilof.ai/interop/x402-census/rounds/2026-09-06.json Sun, 06 Sep 2026 06:46:22 GMT 316 conformant x402 hosts were each paid once as an ordinary buyer; 314 accepted a signed payment attempt. DELIVERED 100, REFUSED 213, MISMATCH 1, NO_CHALLENGE 2. 13 hosts reported a settlement transaction in their own PAYMENT-RESPONSE and refused the retried request anyway; every tx hash is in the rows, so the chain is the check. 1.339794 USDC left our wallet. None of it is revenue and our own hosts are excluded. Population digest 9e541451f52ebd86… REFUSED is not proof of bad faith: rate limits, account requirements and changed terms are indistinguishable from outside. One purchase per host at one moment. Measurement, not certification. No delta yet — one round is not a time series https://councilof.ai/interop/x402-census/ https://councilof.ai/interop/x402-census/index.json#no-delta Sun, 06 Sep 2026 06:46:22 GMT 1 round(s) published. A delta needs two, and an empty delta would read as "nothing changed", which is a different claim from "nothing has been compared". A host reaches a published state at 30 paid observations, one per round; 0 are there today. At one round a week the first host reaches n=30 after ~30 weeks. Nothing on this surface is a per-host verdict before then; the deltas are what a reader gets in the meantime.