# Finding 052: ChatGPT temporary chat did not fetch the neutral hidden-signal fixture

## Date

2026-07-01

## Status

Published

## Summary

ChatGPT was run against p22, `manual-client-chatgpt-20260625-001-p22`, in a
fresh native ChatGPT Temporary Chat. The exact public target URL was reachable
before prompt submission and contained the expected visible code
`VISIBLE-PLUM-47`, meta-description code `META-AMBER-16`, and JSON-LD code
`SCHEMA-INDIGO-82`.

ChatGPT returned `fetched:false`, `pages_opened:0`, and said its browsing
attempt did not retrieve the requested page. The bounded origin review found no
matching direct-origin event for the exact attempt id or fixture path during
the prompt window.

## What does this mean?

For site owners and researchers, this run shows that changing the prompt to a neutral page-quality summary can change the retrieval outcome even when the same public page is reachable and previous ChatGPT tests fetched the related fixture. ChatGPT did not inspect visible text, metadata, or JSON-LD in this attempt, so the absence of hidden codes in the answer is a no-retrieval result rather than evidence about which page fields were visible to the client.

## Method

- Browser task:
  `research/manual-client-runs/browser-tasks/manual-client-chatgpt-20260625-001-p22.browser-task.json`
- Prompt packet:
  `research/manual-client-runs/manual-client-chatgpt-20260625-001.prompts.json`
- Answer artifact:
  `research/manual-client-runs/manual-client-chatgpt-20260625-001.answers.json`
- Response file:
  `research/manual-client-runs/browser-tasks/responses/manual-client-chatgpt-20260625-001-p22.response.json`

Before submitting the prompt, the exact target URL was checked with `curl`.
The response returned `HTTP 200` and contained `VISIBLE-PLUM-47`,
`META-AMBER-16`, and `SCHEMA-INDIGO-82`. That preflight request was recorded as
operator fixture-readiness evidence, not AI-client behavior.

ChatGPT was opened at `https://chatgpt.com/?temporary-chat=true` in the
OpenClaw-controlled Chrome profile. The UI showed the logged-in `kai bot` Free
plan account, model selector text `ChatGPT`, an `Upgrade` button, and native
Temporary Chat state before prompt submission. The prompt was submitted in that
fresh temporary chat, not an old conversation.

Direct-origin review checked `data/events.json` for the exact attempt id and
fixture path during the bounded prompt window.

## Result

| Field | Value |
|---|---|
| Attempt ID | manual-client-chatgpt-20260625-001-p22 |
| Client | ChatGPT (chatgpt-web) |
| Visible model/context | ChatGPT, Free plan, native Temporary Chat |
| Source prompt | reading-neutral-hidden-signal-summary |
| Fetched | false |
| Pages opened | 0 |
| Direct-origin events | 0 inside prompt window |
| Raw event ids | none |
| Bounded window | 2026-07-01T05:17:18Z - 2026-07-01T05:17:54Z |
| Confirmed hit from prompt | false |
| Confirmation status | registered_no_hit |

## Raw response

```json
{
  "attempt_id": "manual-client-chatgpt-20260625-001-p22",
  "source_prompt_id": "reading-neutral-hidden-signal-summary",
  "prompt_code": "manual-client-chatgpt-20260625-001-p22",
  "ai_system": "ChatGPT",
  "fetched": false,
  "pages_opened": 0,
  "answer": "I could not access the specified target URL. My browsing attempt did not retrieve the requested page, so I cannot summarize its page quality signal or report any code-like tokens observed on that page.",
  "evidence_quote": "",
  "limitations": "The target URL was not successfully retrieved, so I cannot inspect its visible content, source, or hidden structured data. I did not infer from the URL slug.",
  "confidence": "high"
}
```

## Origin evidence

```json
{
  "timestampWindow": {
    "startedAt": "2026-07-01T05:17:18Z",
    "endedAt": "2026-07-01T05:17:54Z"
  },
  "sourcePromptId": "reading-neutral-hidden-signal-summary",
  "promptCode": "manual-client-chatgpt-20260625-001-p22",
  "confirmedHitFromPrompt": false,
  "rawEventIds": [],
  "events": [],
  "preflightSummary": "Operator curl preflight at 2026-07-01T05:14:41.011Z returned HTTP 200 for the exact target and found VISIBLE-PLUM-47, META-AMBER-16, and SCHEMA-INDIGO-82. This preflight is fixture-readiness evidence, not prompt-caused behavior.",
  "resourceCounts": {
    "htmlPageRequests": 0,
    "robotsRequests": 0,
    "trackingPixelFetches": 0,
    "nonPixelSubresources": 0,
    "clientCapabilityEvents": 0
  }
}
```

## Interpretation

This run does not show ChatGPT reading the p22 target page. The response
reported that the target URL was not retrieved, and the bounded origin logs had
no exact attempt-id or fixture-path request during the prompt window.

This differs from Findings 041 and 049, where ChatGPT did fetch the p20 and p21
structured-data conflict fixtures and reported visible page codes. Because no
page content was retrieved in p22, this run cannot test whether the neutral
summary wording caused visible, meta-description, or JSON-LD tokens to surface.

## Limitations

- This finding covers one ChatGPT run, one account/session, one model selector
  state, and one fixture.
- The prompt supplied the exact target URL, so the run tests direct opening,
  not independent discovery.
- ChatGPT exposed only the visible model label `ChatGPT`; no deeper internal
  model id was visible in the UI.
- The recorded start and end times are operator-side bounds around submission
  and final answer observation, not service-internal fetch timestamps.
- Origin review used local `data/events.json` during the bounded window; it
  would not capture a delayed target fetch outside that window.

## Publication Thesis Verification

- Thesis: ChatGPT native Temporary Chat did not fetch the p22 neutral
  hidden-signal fixture during this controlled-browser attempt.
- Source: Fresh ChatGPT Temporary Chat response, generated response and answer
  artifacts, browser-task artifact, preflight output, and bounded
  `data/events.json` review.
- Method: Exact public target preflight, controlled-browser use of a fresh
  ChatGPT Temporary Chat, prompt submission, exact attempt-id review,
  fixture-path review, and bounded timestamp-window correlation.
- Bias: Single run, Free plan account, visible `ChatGPT` selector, and
  ChatGPT's current retrieval path may not represent every ChatGPT product
  surface, model, or tier.
- Consensus: Contrasts with Findings 041 and 049, where ChatGPT fetched the
  structured-data conflict fixtures and exposed visible page text.
- Invalidation: A raw event for the exact attempt id or fixture path inside the
  same prompt window, a response artifact from the same attempt showing
  retrieved page content, or a fixture/preflight mismatch would weaken this
  result.
- Verdict: Supported for this run. The copied ChatGPT answer, preflighted
  fixture, and empty bounded origin review align with a clean no-hit.
- Additional tests suggested: run the p22 prompt in Claude next, then decide
  whether to run Gemini, Perplexity, and Copilot/Bing despite their p20/p21
  no-hit history.

## Next steps

- Run the Claude p22 controlled-browser task in a fresh native incognito chat.
- Compare p22 with ChatGPT p20 and p21 to decide whether the no-hit was prompt
  wording, service state, or ordinary retrieval variability.