# Can ChatGPT Sites Show Up in Google? What site:chatgpt.site Reveals

> A Google site:chatgpt.site search reveals public ChatGPT Sites. Learn what is indexed, how to restrict a Site, and when to use protected client sharing.

TL;DR: Public ChatGPT Sites can appear in Google. A search for `site:chatgpt.site` currently returns many individual sites, so a public ChatGPT Site is not a safe place for confidential client files. ChatGPT also offers private and limited sharing modes, but “Anyone with the link” is still not the same as password or verified-email protection.

This is a companion to our guide on [ChatGPT shared links and Google indexing](https://revdoku.com/blog/chatgpt-shared-links-google-indexing-delete-yours/). Shared conversations and ChatGPT Sites use different URLs, but the practical privacy rule is the same: if a page is public, assume it can be found, copied, and retained.

## What `site:chatgpt.site` shows in Google

On August 1, 2026, we searched Google for:

```text
site:chatgpt.site
```

The first page contained multiple individual ChatGPT Sites. Google also offered page after page of additional results.

![Google results for site:chatgpt.site showing multiple indexed sites, with unique subdomains and the signed-in account covered in light gray](/assets/01-google-site-chatgpt-site-results-evidence.webp)

The screenshot keeps the official ChatGPT Sites result visible and masks only the unique subdomains and the signed-in account. Titles, dates, and snippets remain visible so you can see that Google has indexed several distinct ChatGPT Sites. The point is not to expose anyone’s work. The point is that Google has already discovered and indexed public pages on the domain.

The `site:` operator is useful evidence, but it is not a complete inventory. It shows pages Google currently returns for that domain. It does not prove that every ChatGPT Site is public, and it does not list every URL Google knows about.

## Useful searches for finding public ChatGPT Sites

Start with the domain:

```text
site:chatgpt.site
```

Add a word or phrase to narrow the results:

```text
site:chatgpt.site "proposal"
site:chatgpt.site "client"
site:chatgpt.site intitle:"report"
```

To focus on recent results, add a date:

```text
site:chatgpt.site after:2026-07-26
```

Change the date and keywords for your audit. Results vary by time, location, language, and Google’s current index.

Do not use this search to browse other people’s private-looking material. Use it to understand the exposure model and to look for your own public content.

## Not every ChatGPT Site is public

ChatGPT Sites has access controls. In the personal account we inspected, the Share dialog offered **Only you** and **Anyone with the link**.

![ChatGPT Sites Share dialog showing Only you and Anyone with the link, with account and site details covered in light gray](/assets/02-chatgpt-sites-access-options.webp)

OpenAI’s [ChatGPT Sites documentation](https://help.openai.com/en/articles/20001339) describes additional options that can depend on the account and workspace, including selected users or groups, workspace access, and **Anyone on the internet** when public publishing is enabled.

That distinction matters:

- **Only you** is private to the owner and permitted workspace administrators.
- **Selected users or workspace access** can provide a narrower audience when the account supports it.
- **Anyone with the link** means possession of the URL can be enough to open the page. The link can be forwarded or pasted elsewhere.
- **Anyone on the internet** is public publishing. Public pages can be crawled and indexed.

A link that is difficult to guess is not an access-control policy. It is a bearer link. Once it is copied into a message, document, issue tracker, analytics system, browser history, or public page, control over who receives it becomes weaker.

## Why a public ChatGPT Site is a poor client document portal

ChatGPT Sites is useful for public websites, lightweight apps, demos, and experiments. The risk appears when a public site is used as though it were a private delivery room.

Do not place these on a publicly reachable `chatgpt.site` URL:

- confidential proposals or pricing documents
- client reports and presentations
- prototypes containing private customer or company data
- contracts, financial documents, or internal PDFs
- folders of client deliverables
- files that should be visible only to named recipients

Search engines can fetch public pages and create titles and snippets. Other crawlers, monitoring tools, link previews, and archive services may also request public URLs. Some may retain a copy or reference. This does not mean every crawler saves every page, but it does mean that unpublishing later cannot recall every copy already made.

OpenAI also tells Site owners to review a Site before public publishing and avoid sharing confidential, sensitive, or third-party content without the appropriate rights. Its [ChatGPT Sites Terms](https://openai.com/policies/chatgpt-sites-terms/) explain that published content is hosted and made available according to the sharing settings you select.

## How to review and restrict a ChatGPT Site

Use ChatGPT’s own Sites inventory instead of relying on Google:

1. Open **ChatGPT**.
2. Open **Sites** from the sidebar or go to your Sites list.
3. Find the Site and select **Share**.
4. Change **Who has access** to **Only you** or the narrowest audience your account supports.
5. Open **Site settings** if you need more access or publishing controls.
6. If the Site should no longer exist, delete it and confirm the old address no longer loads.
7. Test the old URL in a private browser window.

OpenAI’s official guide says you can restrict access through Share settings or permanently delete a Site. Deletion cannot be undone, so use a narrower access setting when you only need to stop public access temporarily.

## Removing the page does not instantly erase Google’s result

After a page is restricted or deleted, Google may still show an old title or snippet until it recrawls the URL.

Google’s [Refresh Outdated Content](https://support.google.com/websearch/answer/6349986?hl=en) guidance explains that deleted or changed pages can temporarily remain in search results with outdated descriptions or cached information. Its [removal guide](https://support.google.com/websearch/answer/11080680?hl=en) recommends removing content at the source first, then using the appropriate refresh or removal tool.

A practical cleanup sequence is:

1. Restrict or delete the ChatGPT Site.
2. Confirm the old URL no longer exposes the content.
3. Search Google for the exact URL and distinctive phrases.
4. Request an outdated-content refresh if an old result remains.
5. Assume any copy already downloaded by another person is outside your control.

## Use Revdoku for protected client sharing

If the deliverable is intended for a client rather than the public web, publish it behind an actual access boundary.

Revdoku can host an AI-built website, exported HTML project, PDF, presentation, report, or folder with:

- password protection
- verified-email access
- email and domain allowlists
- visitor notifications and activity
- file updates without changing the shared URL

![Revdoku password access settings, with the private URL and password covered in light gray](/assets/03-revdoku-protected-client-sharing.webp)

Protected content is not available to Google as a public page. You can give a client one stable link, update the files behind it, and keep the audience limited.

## Try Revdoku for free

Use ChatGPT Sites for public work you want people to find. Use Revdoku when a specific client or approved audience should be able to open the result.

[Create a protected Revdoku link](https://app.revdoku.com/users/sign_up) or read the guide to [password-protecting a website](https://revdoku.com/password-protected-website/).

Start publishing for free

Connect this AI to Revdoku so it can store files, create, update, and publish websites for me. Read and follow https://revdoku.com/prompt.md. A Revdoku account is required. If I do not have one, send me https://app.revdoku.com/users/sign_up?utm_source=revdoku.com&utm_medium=ai-chat&utm_campaign=connect_ai_first and wait until I confirm signup is complete. For ChatGPT, Claude, or another cloud AI, connect https://app.revdoku.com/mcp and complete Revdoku OAuth. For a local agent, use Revdoku's browser device sign-in and complete the connection in Revdoku. Never ask me for a password or API key in chat, or for a verification code. When Revdoku tools are connected, verify the connection, tell me you’re ready, and ask what I want to publish.

Copy Prompt

or

Create Free Account

## Frequently asked questions

### Are all ChatGPT Sites indexed by Google?

No. A Google `site:chatgpt.site` search proves that some public ChatGPT Sites are indexed. It does not show every Site, and private Sites should not be treated as public pages.

### Is “Anyone with the link” private?

It is narrower than public discovery, but it is not the same as a password or verified identity check. Anyone who receives the URL may be able to open and forward it.

### Can Google keep showing a Site after I delete it?

Yes, temporarily. An outdated title or snippet can remain until Google recrawls the URL. Remove or restrict the source first, then use Google’s outdated-content tools if needed.

### Should I publish client PDFs or reports on a public ChatGPT Site?

No. A public website is the wrong container for confidential client material. Use a protected sharing layer with an access gate and a controlled audience.

### Can I update a file in Revdoku without sending a new link?

Yes. You can replace or update the PDF, website, or other files behind the existing Revdoku URL, so the client keeps using the same link.

---

[View the canonical page](https://revdoku.com/blog/can-chatgpt-sites-show-up-in-google-what-site-chatgpt-site-r/) · [Browse llms.txt](https://revdoku.com/llms.txt)
