What was read, and when.
Statements about AI engines go out of date. This log records the date on which each provider's documentation was read for this site, what it said then, and every change to the method.
This change log is a dated record of two things: what the documentation of AI providers said on the day it was read for this site, and what changed in the measurement method. It exists so that statements about provider access controls, and the studies quoted most often on the site, can be traced to a date. It records documents read, not behavior observed: no engine test is reported here. The first entry is the baseline of October 4, 2026.
How entries are written
Provider documentation is re-checked every quarter, and each check adds an entry. Each entry gives a date, what was read or changed, and where the site reflects it. An entry is never edited after publication; a correction is a new entry that refers to the earlier one. Entries about engines describe documentation. They do not claim that an engine behaves as its documentation says.
October 4, 2026: baseline
Provider documentation read
| Provider | What the documentation said on that date | Reflected in |
|---|---|---|
| No additional requirements or special markup to appear in AI Overviews or AI Mode; a supporting link must be a page indexed and eligible for a snippet in Search | Google AI Overviews, Google AI Mode | |
Google-Extended is a control separate from those of Google Search |
Gemini | |
| OpenAI | Three user agents: OAI-SearchBot for search, GPTBot for training, ChatGPT-User for user-triggered actions, for which robots.txt rules may not apply; sites opted out of OAI-SearchBot are not shown in ChatGPT search answers, though the link and title of a disallowed page may still be surfaced if obtained from a third-party search provider |
ChatGPT |
| Perplexity | Two agents, PerplexityBot and Perplexity-User; the second "generally ignores robots.txt rules"; IP ranges are published |
Perplexity |
| Anthropic | Three agents, ClaudeBot, Claude-User and Claude-SearchBot; page last updated by Anthropic on April 7, 2026 |
Claude |
| Microsoft | Only a Bing Webmaster Blog post of September 22, 2023, about Bing Chat, could be read: NOCACHE and NOARCHIVE. Bing's current help pages could not be read |
No page: the Copilot page is deferred until current documentation is available |
The sources are linked from each platform page. The Microsoft entry rests on one document from 2023, and whether it applies to Copilot was not verified; for that reason no Copilot page is published.
Third-party studies read
| Study | Date given by the source | Used in, among other pages |
|---|---|---|
| Status Labs, white paper on AI and reputation management | 2026 | How sources shape representation |
| Peec AI, analysis of five million query fan-outs | April 2026 | Why rankings do not guarantee citations |
| LLM Pulse, measuring AI share of voice | Published June 1, 2026; updated September 28, 2026 | AI share of voice |
| Aggarwal et al., "GEO: Generative Engine Optimization" | Submitted November 2023; last revised June 2024; KDD 2024 | Generative engine optimization |
Method
- Protocol published. The methodology is stated before any result.
- Prompt library, version 1. Thirteen numbered templates in two groups, as listed in the prompt library.
- No blended score. Any result is to be reported per engine and per observation.
- No measurement made or published. As of this entry, the site reports no test of any engine.
What this log does not cover
It does not track every change made by every provider. It records what was read for this site, on the dates given. A provider may have changed its documentation since the last entry; the date of the last check is the limit of what the site can state.
It does not record changes in how engines answer. That would be a measurement, and it would be published under the methodology.
Where it sits
The log is part of Research. Monitoring over time for one entity is the engagement described in AI Reputation Intelligence, and corresponds to the last layer of The Authority Architecture System, Monitoring and Governance.
Understand how search and AI currently interpret your authority.
Request a confidential assessment of your visibility, entity signals, narrative consistency and source authority.
Request a Private AssessmentSelective engagements for leaders and organizations with material reputational stakes.