ChatGPT Locked Chats: what the PIN hides (and doesn't)

Short answer: ChatGPT's new Locked Chats feature hides chosen conversations behind a 4-digit account PIN (or Face ID, Touch ID or device unlock on supported phones). It keeps them out of your regular history, pinned chats and search, which helps if someone borrows your phone or sees your screen. It does not encrypt them end to end, change how long OpenAI keeps them, or change whether they can be used to improve its models. And if you forget the PIN and reset it, every locked chat is deleted for good.

What happened

On October 20, 2026, OpenAI began rolling out Locked Chats to ChatGPT Free, Go, Plus, Pro, Business, Enterprise and Edu plans. The rollout is gradual. If you do not see Lock chat in the ••• menu of an eligible chat, it may not have reached your account yet.

OpenAI says the goal is to keep sensitive conversations out of view when you share your screen or use a shared device. Fortune and TechRadar both covered the launch and make the same point: the feature controls who can open a conversation, not how the company handles it.

How Locked Chats works, step by step

1. Open the ••• menu for a chat, select Lock chat and confirm.
2. If you have not set one yet, create a 4-digit account PIN. The same PIN works on your other devices when you are signed in to the same account.
3. The chat moves to Projects → Locked chats. Its title and content disappear from regular history and regular search.
4. To open it later, go to Projects → Locked chats and enter your PIN, or use Face ID, Touch ID or device unlock where supported. This gives you temporary access and does not return the chat to regular history.
5. When you are done, select Lock now. Otherwise access ends on its own: on the web when you refresh or close the tab, and on phones after about a minute away or in the background.

To move a chat back, open it, choose Unlock chat and confirm. It returns to regular history, and if your settings allow, the full conversation may then be used for regular Memory.

What Locked Chats does not do

OpenAI's own help page is unusually clear here. Locking a chat changes where it appears and how you open it. It does not:

• create end-to-end encryption or a separate encrypted vault
• change how long OpenAI retains the chat
• change your "Improve the model for everyone" setting
• remove information that ChatGPT remembered before you locked the chat
• override workspace policies, safety requirements, or legal and compliance obligations

In other words, it protects you from people around you, not from the service itself.

The fine print worth knowing

Forgetting your PIN is costly. Changing a PIN you remember keeps your chats. Resetting a forgotten PIN permanently deletes all locked chats, memories stored only inside them, and files or other content protected only by the old PIN. OpenAI says this cannot be undone and support cannot recover them.

Memory is only partly separated. Locked chats can use relevant information from your regular Memory. Things ChatGPT remembered before you locked a chat stay in your regular memory until you delete them yourself. Details from messages sent after locking can be remembered for other locked chats, but do not carry into regular chats. If you know a conversation will be sensitive, lock it before you start, not halfway through.

Shared links keep working. Locking a chat does not delete a link you already shared. People who have it can still see what was shared before, although new messages sent while the chat is locked are not added. Delete the link separately.

Access does not always end right away. On the web, the chat can stay open in the tab where you locked it, and other devices where ChatGPT is already open may keep showing its title until they refresh. Check other open devices before you share your screen.

Locked is not copy-proof. Anyone who can open a locked chat can still copy its contents. Files uploaded or created in the chat are protected, but files added from elsewhere in Space stay available there.

Not every chat can be locked. Temporary chats, group chats, archived chats, chats created from a shared link, and chats with scheduled tasks or meeting recordings are among the exceptions. The feature is also not available on ChatGPT for Healthcare.

Workplace accounts are different. In Enterprise and Edu workspaces, locked chats are still included in the logs available through OpenAI's Compliance Platform. Administrators cannot unlock your chats or reset your PIN, but your organization's policies still apply.

Locked Chats or Temporary Chat?

They solve different problems:

• Locked Chats keep a conversation saved so you can come back to it, but hidden behind a PIN. They follow your existing model-improvement setting.
• Temporary Chat is for conversations you do not need to keep. OpenAI says temporary chats do not appear in history, do not create or update memories and are not used to improve its models while they remain temporary, though a copy may be kept for up to 30 days for safety reasons.

If you save a temporary chat, it becomes a regular chat and follows your normal settings. To control whether your chats help improve the model, review your data controls separately.

What to do today

1. Decide which conversations deserve a lock (health, money, relationships, surprise plans) and lock them before you type the sensitive part.
2. Keep your PIN somewhere safe, such as a password manager. A forgotten PIN means a reset, and a reset means deletion.
3. Delete shared links to any chat you lock.
4. Review what Memory has saved and delete anything you would rather ChatGPT forget.
5. Check Settings → Data Controls and decide whether you want your chats used to improve the model.
6. Tap Lock now when you finish, and refresh or close other tabs before handing over your screen.
7. Keep passwords, ID numbers and confidential work documents out of any chatbot, locked or not.

What this means for everyday users

For the classic situation (a partner, child or coworker picking up your phone or glancing at your screen) Locked Chats is a useful extra layer. TechRadar's example is planning a surprise gift you do not want to pop up when someone borrows your device. For real privacy from the company, the old advice still applies.

Fortune also notes the bigger picture: a Stanford study from October 2025 examined the privacy policies of six leading U.S. AI developers and found all six appeared to use chat data to train their models by default, although practices and opt-out options varied. A PIN that hides a chat from people nearby is welcome, but it is worth knowing exactly where that protection stops.

Frequently asked questions

Are locked chats encrypted? OpenAI says Locked Chats does not create end-to-end encryption or a separate encrypted vault.

Can my locked chats still be used to train the AI? They follow your existing "Improve the model for everyone" setting. Locking does not change it.

I forgot my PIN. Can I get my chats back? Not after a reset. A reset permanently deletes locked chats and the content protected only by the old PIN, and OpenAI Support cannot recover it.

Can my employer or school see my locked chats? Administrators cannot unlock them or reset your PIN, but in Enterprise and Edu workspaces locked chats are still included in logs available through OpenAI's Compliance Platform.

This article reflects OpenAI's published information and news coverage as of October 25, 2026. The feature is still rolling out, so check OpenAI's help center for the latest details.

Related reading:

• Is ChatGPT safe? What happens to what you type
• How to delete or export your AI chat history
• Why does AI forget what you told it?