30 ChatGPT Sites Entered Google Search. Then 20 Experienced AI Index Drop and 10 Experienced AI Index Ghosting
- AI Trending Group

- Sep 1
- 9 min read
What happens after an AI-generated website is discovered and indexed by Google?
An independent SEO case study involving 30 AI-generated websites hosted on ChatGPT subdomains revealed a search visibility cycle with four proposed stages:
AI Index Entry
AI Index Ascent
AI Index Drop
AI Index Ghosting
The experiment was observed and documented from July 25 through August 29, 2026.
All 30 AI websites initially appeared in Google Search. Some gained surprisingly strong early visibility, including one Spanish written AI website that reached the first position on page two for its monitored search query:
At first, the experiment appeared successful. Google had discovered the websites, processed their pages, and allowed them to compete in search results as seen below:

However, that visibility did not last.
As the websites were monitored over the following days, 20 disappeared from the Google searches used to locate them, including repeated site: searches that had previously returned the websites.
The remaining 10 followed a different path. They continued to appear when checked with the site: command, but their keyword positions progressively declined. Some moved from page two toward page six and continued falling until they became practically unreachable through ordinary searches.
These two outcomes are described as AI Index Drop and AI Index Ghosting.
Purpose of the Case Study
The purpose of the experiment was to observe how Google and other search engines treated newly published AI-generated websites hosted on ChatGPT subdomains.
The investigation focused on several questions:
* Would Google discover and index these AI websites?
* How quickly would they appear in search results?
* What initial positions would they receive?
* Would their visibility increase after being indexed?
* Would they maintain their initial positions?
* Would they remain discoverable after several days?
* Would their visibility improve, stabilize, decline, or disappear?
The experiment also explored a larger question being discussed within the SEO community: Could certain AI-generated websites or AI-hosted subdomains receive reduced visibility after initially appearing in Google?
The goal was not to assume the answer. The goal was to document the sequence of events after the websites were published, submitted, indexed, and monitored.
Stage 1: AI Index Entry

The first stage observed in the experiment is called AI Index Entry.
AI Index Entry is a proposed SEO term describing the moment when an AI-generated website, webpage, or AI-hosted subdomain is discovered, crawled, processed, and initially added to a search engine’s searchable index.
This includes AI-generated websites published on custom domains as well as websites hosted on AI platform subdomains, such as addresses ending in chatgpt.site.
One example would be: Open AI SEO
All 30 ChatGPT websites examined in this case study reached the AI Index Entry stage.
They were initially discoverable through Google Search, including searches performed with the site: command. This demonstrated that Google was capable of finding, processing, and displaying AI-generated websites hosted on ChatGPT subdomains.
AI Index Entry is important because it separates website publication from actual search engine discovery.
Publishing an AI-generated website or launching an AI-hosted subdomain does not automatically mean it has entered Google’s index. The website must first be discovered, crawled, and processed before it has an opportunity to appear in search results.
However, entering the index does not guarantee strong rankings, meaningful traffic, or permanent visibility. It only represents the beginning of the AI website’s search visibility journey.
Stage 2: AI Index Ascent

After AI Index Entry, several websites entered a period of improving visibility called AI Index Ascent.
AI Index Ascent is a proposed SEO term describing the period when an AI-generated website, webpage, or AI-hosted subdomain begins gaining search visibility after being indexed.
During this stage, a website may:
* Appear for more search queries
* Gain stronger keyword positions
* Receive additional search impressions
* Have more pages discovered or indexed
* Move higher in search results
* Reach its strongest observed search position
One website in the experiment reached the first position on page two for its monitored search query.
For a completely new website hosted on a new ChatGPT subdomain, this was a promising early result. It showed that an AI-generated website could enter Google’s index and begin competing for meaningful search positions.
This early rise is important because the websites did not immediately disappear after publication. They first entered Google, gained visibility, and established measurable positions.
The decline happened afterward.
Stage 3: The Visibility Peak and Mount AI
Following AI Index Entry and AI Index Ascent, the websites reached their highest observed level of search visibility.
This peak resembles the concept known as Mount AI.
Mount AI generally describes a pattern in which AI-generated or machine-scaled content gains increasing organic visibility, reaches a peak, and then begins declining.
The term Mount AI was coined by SEO consultant Glenn Gabe. It was not created as part of this case study and should not be attributed to Esteban Maraboli.
Mount AI is also not being presented as one of the four proposed stages developed through this case study. It represents the peak or tipping point between AI Index Ascent and the two possible outcomes observed afterward.
The ChatGPT websites did not necessarily generate a massive mountain of organic traffic. However, their visibility followed a similar pattern:
1. The websites were published.
2. Google discovered and processed them.
3. They completed AI Index Entry.
4. Some gained stronger positions through AI Index Ascent.
5. They reached their highest observed visibility.
6. Their search positions began declining.
7. They experienced either AI Index Drop or AI Index Ghosting.
Both declining patterns appeared after the initial visibility peak.
Learn more about Glenn Gabe’s discovery and read his original post on X here: Mount Ai
Stage 4: AI Index Drop

Twenty of the 30 ChatGPT subdomains appeared to experience AI Index Drop.
These websites were initially discoverable through the searches used during the experiment. Later, they could no longer be found through those same searches, including repeated site: checks that had previously returned the websites.
They were not simply observed a few positions lower. They became undiscoverable through the methods used to monitor them.
AI Index Drop is a proposed SEO term describing an AI-generated website, webpage, or AI-hosted subdomain that initially appears in Google’s searchable index but later becomes absent from the searches used to locate it.
In this experiment:
* 20 of 30 websites entered the observed AI Index Drop group
* This represented approximately 66.7% of the sample
* Each website initially appeared in Google
* Each website later became absent from repeated searches used to check it
The findings suggest that Google may have removed, suppressed, reevaluated, or otherwise stopped displaying these websites in its searchable results.
However, the experiment does not establish the exact technical mechanism behind their disappearance. A failed site: search is meaningful observational evidence, but the site: command is not a complete diagnostic tool or a guaranteed measurement of Google’s entire index.
The study also does not prove that Google applied an intentional penalty. Other possible factors may include crawling changes, canonicalization, content evaluation, duplication, technical signals, hosting characteristics, quality reassessment, or broader indexing processes.
Based on the repeated observations, these 20 websites are classified as apparent AI Index Drop cases within this study.
Why AI Index Drop Matters
AI Index Drop gives SEO professionals, researchers, and website owners a more precise term for describing the disappearance of an AI-generated site after it has already entered a search engine’s index.
Without this term, several different situations might be described simply as “not indexed,” even though they represent very different events.
Stage 5: AI Index Ghosting

The remaining 10 websites followed a different path.
These sites continued to appear in Google when checked with the site: command for more than a week. They avoided the apparent indexing disappearance experienced by the other 20 websites.
However, their rankings for their target keywords progressively deteriorated.
Pages that initially appeared around page two moved toward page six and continued falling deeper into the results. Although the websites remained present in Google’s searchable index, they became almost invisible during ordinary keyword searches.
This pattern is called AI Index Ghosting.
AI Index Ghosting is a proposed SEO phenomenon in which an AI-generated website, webpage, or website hosted on an AI platform subdomain remains present in Google’s index and continues appearing through the site: command, but becomes almost invisible during manual searches for its target keywords after progressively losing its ranking positions.
The word “ghosting” describes what happens from the searcher’s perspective.
The website is still there, but ordinary users are extremely unlikely to see or visit it through traditional organic search.
From a technical perspective, the page remains discoverable.
From the searcher’s perspective, it has become a ghost page.
In this experiment:
* 10 of 30 websites experienced AI Index Ghosting
* This represented approximately 33.3% of the sample
* The websites remained discoverable through site: searches
* Their keyword rankings progressively deteriorated
* Some fell from page two toward page six or deeper
* They became practically unreachable through ordinary searches
AI Index Drop vs. AI Index Ghosting
The difference between these two proposed terms depends on what happens after the website’s initial visibility period.
AI Index Drop
The website initially appears in Google but later becomes absent from the searches used to locate it, including repeated site: searches that previously returned the website.
The page appears to lose its searchable presence.
AI Index Ghosting
The website remains discoverable through the site: command, but its target keyword rankings decline so severely that ordinary users are extremely unlikely to encounter it.
The page remains searchable in a technical sense, but loses practical organic visibility.
The simplest distinction is:
AI Index Drop means the website appears to leave the searchable index.
AI Index Ghosting means the website remains in the searchable index but becomes a ghost page in organic search.
How Is AI Index Ghosting Different From AI Ghosting?
AI Index Ghosting should not be confused with the similar term AI Ghosting.
Alexandre Rastello, founder and CEO of Mentionable, uses AI Ghosting to describe a situation in which AI tools ignore a brand in their answers or recommendations, even when the brand should logically appear.
That concept concerns brand visibility inside AI tools such as ChatGPT, Claude, Gemini, and Perplexity.
AI Index Ghosting concerns visibility inside traditional Google Search.
A website experiencing AI Index Ghosting can still appear through the site: command, but becomes extremely difficult to find when users manually search for its intended keywords.
The two concepts involve different visibility environments and should be treated as separate terms.
Why Does This Visibility Cycle Make Sense?
Initial indexing is not necessarily Google’s final evaluation of a new website.
When Google first discovers a page, its systems may still need time to:
* Crawl and process the complete website
* Select the preferred canonical URL
* Compare the content with competing pages
* Evaluate relevance for different searches
* Understand the website’s purpose and subject
* Assess originality, quality, and usefulness
* Gather authority and trust signals
* Evaluate the hosting environment and technical structure
* Recalculate rankings as additional information becomes available
A newly indexed page may receive temporary visibility before Google determines where it should rank over a longer period.
In this case study, the sequence was consistent across the sample:
AI Index Entry came first.
AI Index Ascent followed for websites that gained stronger visibility.
The websites then reached their highest observed positions.
AI Index Drop or AI Index Ghosting followed afterward.
This sequence suggests that initial indexing was not the final challenge. Maintaining meaningful search visibility became the more significant problem.
Questions About Shared Platform Signals
All 30 websites were AI-generated and hosted on ChatGPT subdomains with addresses ending in chatgpt.site.
This raises questions about whether shared platform characteristics may have influenced Google’s evaluation.
The websites may have shared:
* Similar technical structures
* Related hosting characteristics
* Common subdomain patterns
* Limited domain history
* Similar publishing environments
* Other platform-level signals
The observed results are consistent with the possibility that certain AI-generated websites, especially those hosted on AI platform subdomains, may receive reduced visibility after an initial indexing period.
The experiment does not prove that Google intentionally penalized the websites because they were AI-generated. It does, however, establish a measurable pattern that deserves additional testing with larger samples, longer monitoring periods, custom domains, different content types, and comparison groups.
Final Findings
Out of 30 AI-generated websites hosted on ChatGPT subdomains:
* All 30 completed AI Index Entry and initially appeared in Google Search
* Several entered AI Index Ascent and gained promising early positions
* One reached the first position on page two for its monitored query
* All eventually experienced a decline from their highest observed visibility
* 20 became undiscoverable through the searches used in the experiment
* Those 20 represented approximately 66.7% of the sample
* The 20 sites were classified as apparent AI Index Drop cases
* 10 remained discoverable through the site: command
* Those 10 became almost invisible in manual target keyword searches
* The 10 sites represented approximately 33.3% of the sample
* Those 10 followed the proposed AI Index Ghosting pattern
* Both outcomes appeared after a visibility peak resembling Mount AI
The case study suggests that getting an AI-generated website or AI-hosted subdomain indexed may only be the first challenge.
The larger question is whether that website can maintain meaningful organic visibility after Google has had more time to crawl, process, compare, and evaluate it.
Attribution
The proposed SEO terms AI Index Entry, AI Index Ascent, AI Index Drop, and AI Index Ghosting, along with the definitions and visibility framework presented in this case study, were introduced or developed by Esteban Maraboli, founder of E&L Interactive Web Design LLC.
The terms emerged from his independent research involving 30 AI-generated websites hosted on ChatGPT subdomains and his observations of Google Search indexing, search visibility, and Google Search Console submission behavior.
The research began with an observed case on July 25, 2026, and expanded through additional experiments conducted during August 2026.
The term Mount AI was coined by SEO consultant Glenn Gabe and is included only to describe the visibility peak that may occur before AI Index Drop or AI Index Ghosting.
These terms represent independent SEO research and proposed terminology. They are not official Google terminology or confirmed descriptions of Google’s indexing systems.






Comments