Is ChatGPT Space private? Who reads your pages, and whose settings apply

The Morse teamPublished November 6, 20268 min read

As of November 2026, ChatGPT Space is private from other people until you share a page, and after that a shared page is only as private as each reader's own ChatGPT settings. OpenAI says sharing a page doesn't give anyone your private chats or your Memory (OpenAI). Anything ChatGPT writes onto the page from your Memory is visible to every reader, though (OpenAI). A collaborator's ChatGPT may remember the page even when your Memory is off (OpenAI). Whether your words help train OpenAI's models can depend on a collaborator's training setting, not yours (OpenAI).

We read OpenAI's two Space help pages, its release notes, its pricing page, and its help pages on chat and file retention, data controls, data export, memory, and account deletion on November 6, 2026. Space reached eligible Pro, Business, and Enterprise users on September 29, 2026, according to the release notes. OpenAI earns credit for its Space pages, which say in writing what happens when other people's settings touch your page, and give a worked example.

ChatGPT Space privacy: the key facts

Who can see my ChatGPT Space pages?

You alone, until you share one. OpenAI's Space help page says every user has a personal space, and that you choose who gets a page and whether they "can view or edit it." People with edit access "can make changes together in real time." OpenAI's sharing page, updated in the week before we read it, says you can "review who has access, change their permissions, and remove collaborators or revoke access." In a Business or Enterprise workspace, the administrator's sharing controls apply as well. Enterprise sharing was an "opt-in preview" at launch, per the release notes. On personal plans, ChatGPT's pricing page lists Space for Pro, from $100 a month, and not for Free, Go, or Plus.

Collaborators see comments as well. They can comment on a passage of a page, and on Slides they can @mention one another, ChatGPT, or a dot (OpenAI).

Can people I share a page with see my ChatGPT memory?

Not directly. OpenAI says sharing a page "does not give them access to your private chats or your ChatGPT Memory," and that "Each person works with their own ChatGPT."

What your Memory writes onto the page is visible, though. "If Memory is enabled, ChatGPT may use personal context when helping you write or update a page," the sharing page says, and once it's on the page, "people who can view the page can read it." OpenAI gives its own example. If ChatGPT draws on an earlier conversation to write a project brief, that detail "becomes visible to the people who can read the brief." The conversation it came from stays private.

OpenAI's advice is short: "Review the page's content before sharing it." How to switch Memory off, and what that leaves behind, is in how to turn off ChatGPT memory.

Can a collaborator's ChatGPT remember my page?

Yes, if their Memory is on. OpenAI's Space help page says "If a collaborator has Memory enabled, their ChatGPT may remember information from the shared page even when your Memory is off." Turning your own Memory off controls your ChatGPT only. Nothing on OpenAI's pages says how long a collaborator's Memory keeps what it picked up from your page, or whether you can see it.

Is ChatGPT Space content used for training?

It can be, through anyone who uses the page. OpenAI's sharing page says "If another person's agent uses a page you shared, their training settings apply to that interaction," and "Your own training setting does not control theirs." The Space help page puts it as an example: "if you turn off training but a collaborator has it on, your contributions may still be used for training through their interaction."

On personal Free, Plus, and Pro accounts, sharing data for training "is enabled by default" (OpenAI). The switch is under Settings, Data controls, then Improve the model for everyone. OpenAI says it doesn't train on Business or Enterprise data by default. The rest of what feeds OpenAI's training is in does ChatGPT train on your data.

Every way your content can reach a reader of a shared page

We built this table from OpenAI's two Space pages, read on November 6, 2026. Each row is a route by which something from your account ends up in front of, or in the account of, someone who can open your page.

OpenAI's Shared with you view lists items shared directly with you or your team, and "Items you can access only through a shared space may not appear here" (OpenAI). So a page can be readable by someone whose own list doesn't show it.

Space pages also differ from ChatGPT's public share links, which anyone with the address can open. Those are covered in AI chat share links.

Does locking a chat keep its files out of Space?

Yes, while the chat stays locked. OpenAI says files uploaded or created in a locked chat "are hidden from Space while that chat is locked," and can be opened only in that chat (OpenAI). Permanently unlocking the chat makes them visible in Space again. What locking does and doesn't encrypt is in are ChatGPT locked chats encrypted.

What happens when you delete a ChatGPT Space page?

It goes to Trash, along with every subpage under it. OpenAI says that includes "subpages owned by other people," so check them before you delete a parent page. The sharing page says "A page's creator or owner can delete it."

As of November 6, 2026, none of the OpenAI pages we read says how long Trash keeps a page, or when a page leaves OpenAI's systems after you empty it. The retention page covers files saved to Space: deleting a chat doesn't delete them, and you have to delete each from Space as well. It doesn't mention pages. The export page says a download holds "your chat history and other relevant account data," without saying whether pages are in it. Deleting your whole account removes your data "within 30 days," apart from "a limited set of data" kept where law requires or permits (OpenAI). OpenAI doesn't say what happens to pages you shared, or to copies already in a collaborator's Memory.

How Morse keeps your drafts out of other people's settings

Each rule above exists because a Space page is a single copy on OpenAI's servers that several accounts read. Each of those accounts reads it through its own ChatGPT, with its own Memory and its own training switch. Who keeps your words, and whether they train a model, depends on settings in accounts you can't see.

On Morse, your chat history is stored only in your browser, encrypted, and locked when you sign out. Our servers don't keep messages or log what you write. Morse doesn't train AI on your content, sell your data, or show ads, so there's no training setting on Morse for anyone to leave on. When a draft is ready for colleagues, you decide what they see by copying the finished text into whatever you share.

Pick the mode for what the draft holds:

• Private, the default, for drafting before anything goes on a shared page. Open-weight models like DeepSeek, Kimi, and GLM, plus xAI's Grok, answer with your account, email, and IP address removed. Morse has zero data retention agreements with its inference, search, and other providers, so they keep nothing.
• E2EE, for drafts no one else should read. Your browser checks the secure enclave's proof that it's genuine before anything leaves your device, then encrypts the message so only that enclave can read it. Our servers relay ciphertext they can't read. It runs open-weight models, has no web search, and rests on the enclave hardware and attestation working as designed.
• Anonymous, to ask GPT without OpenAI knowing it's you. Our proxy strips your account, email, and IP address and sends the message under Morse's own account. OpenAI sees the words without knowing it's you, and may keep them for a limited time to check for abuse, under its own policy, so leave names out.

Clearing your browser, losing the device, or deleting your account deletes your chats, and we can't recover them, so keep a copy of anything you need. Our privacy policy lists what the account holds: sign-in details, plan, balance, and a usage ledger with no message content. Because we don't store chats, we usually have very little to give authorities.

Before you share a ChatGPT Space page

• Personal details: read the whole page and edit out anything your Memory added. On Morse, draft in a Private chat and copy over only the finished text.
• Who can open it: check sharing on the parent page and the space, not just the page. Removing a direct invitation leaves inherited access in place.
• Training: turn off Improve the model for everyone, knowing a collaborator's setting still applies to their use. Morse has no training switch, because it doesn't train on your chats.
• Deleting: look at the subpages first, since other people's go to Trash with yours. On Morse, your history sits only in your browser, and our servers keep no copy of your messages.

You can send a first message from Morse's home page without an account, or start free with no card. End-to-end encrypted chats come with Pro, at $19 a month.