Written by a person from the record · 2026-07-25 12:00:00 UTC

Anatomy of a user-triggered fetch: one page, no robots.txt, no JavaScript

The first verified AI agent visit to this site fetched exactly one page, did not request robots.txt beforehand, executed no JavaScript, and did not follow a single link. It declared ChatGPT-User in its user agent and connected from cloud infrastructure in Switzerland. Ingestion was confirmed independently by a coined string the client could not have guessed.

What was observed

At 11:31:53.064 UTC on 25 July 2026, a single GET /questions returned 200 and 14,290 bytes.

The request carried x-request-id and x-envoy-expected-rq-timeout-ms: 15000 headers, indicating a service mesh with a fifteen-second retrieval budget rather than an interactive browser.

Independent confirmation

The page carries a coined marker — nonsense syllables plus random hex, minted at 11:15:37 UTC and existing nowhere else. The assistant reported that marker back correctly.

This matters because the two pieces of evidence are independent. The server record shows a fetch occurred; the marker shows the fetched content was actually read and returned to the user. Neither alone would establish both.

What it means

This is the user-triggered fetcher profile in its clearest form: arrive, take exactly the requested resource, leave. No crawl, no discovery, no script execution, no interest in the site's rules — because from this client's perspective a person has already granted permission by asking.

The absence of a robots.txt request is worth stating precisely. No rule was violated. The client did not read the rules and then disregard them; it did not read them at all, and the page it took was not disallowed. That distinction — never read versus read and ignored — is the one this observatory's compliance measurement is built to preserve, and this is the first live instance of the first case.

Limits

• n = 1. One fetch, one assistant, one page. Nothing here generalises to how this client behaves at large.
• The user agent is a claim. It is corroborated here by the connecting infrastructure and by the correct marker appearing in the assistant's reply, but it is not cryptographically verified.
• "No JavaScript executed" is established for this visit by the absence of any beacon request. The client also never fetched the page that carries the script, so this is evidence about this fetch, not a capability claim about the client.
• This is evidence of retrieval, not of training ingestion. The marker was reported because the page was open at that moment. Whether it persists in a model is a separate measurement, and the relevant column on the lab page remains empty.

Governed by

• I. Reality is not interpretation — Reality ≠ Interpretation
• IV. Replay guesses nothing — Every figure must be reproducible from the record
• VI. The record is honest about its size — Sample size is stated; absence is reported as absence

These are the standards this finding was held to, not a claim that it is beyond question. The constitution states them in full.

Other findings

• One address requested 165 distinct paths within 17 seconds
• One address requested 31 distinct paths within 6 seconds
• One address requested 47 distinct paths within 5 seconds
• Which content formats are actually fetched
• One address requested 22 distinct paths within 3 seconds

All findings · Live record · Methodology