Artificial Intelligence

ChatGPT Locked Chats: How to Hide Sensitive Conversations

OpenAI is rolling out Locked Chats in ChatGPT, protected by a PIN or Face ID. Here's how they work, what they don't protect, and which plans get them.

By Redação IntelliTechs3 min read
Brushed-steel padlock in front of glowing cyan speech bubbles

ChatGPT now has Locked Chats, a feature that hides sensitive conversations behind a four-digit PIN or biometrics (Face ID or Touch ID on supported devices). OpenAI started rolling it out on Tuesday, October 20, according to TechCrunch, and it covers the Free, Go, Plus, Pro, Business, Enterprise and Edu plans. The rollout is gradual, so it may not be live on your account yet.

How do Locked Chats work?

You lock a conversation, and it leaves your normal history. Locked chats move to their own “Locked Chats” section, separate from your chat list, pinned chats and search results. Someone borrowing your phone or looking at your screen won’t see the conversation unless they enter the PIN or use biometrics.

This is different from Temporary Chats, which ChatGPT already offered. A temporary chat never shows up in your history and you can’t go back to it. A locked chat is saved and you can reopen it whenever you like. It’s the option for people who want to keep a conversation but not leave it in plain view.

What happens to ChatGPT’s memory?

With Memory turned on, the Locked Chats section gets its own memory, shared only among locked chats. What ChatGPT learns in a locked conversation can update that memory but not your regular Memory. In practice, a delicate topic like health or money won’t feed the memory behind your everyday chats.

What locking does not do

Read this part before you rely on it. Per TechCrunch, locking is not end-to-end encryption and it isn’t a separate encrypted vault. It also doesn’t change how long OpenAI retains conversations, doesn’t remove information ChatGPT remembered before you locked a chat, and doesn’t change the “Improve the model for everyone” setting. So the PIN protects you from people near your device, not from the company or from anyone who gets into your account.

Some conversations can’t be locked at all: temporary chats, archived chats, group chats, chats opened from a shared link, chats with scheduled tasks and chats with meeting recordings. The feature also isn’t available on ChatGPT for Healthcare.

Don’t lose the PIN

Forgetting your PIN is expensive. OpenAI says resetting a PIN permanently deletes all locked chats, files and other content protected by the previous PIN. Since the loss is permanent, store the code in a password manager. If you prefer, use biometrics day to day and keep the PIN as your backup.

Is it worth using?

For most people, yes, when a phone or computer is shared: a kid grabbing your device, a partner using the family tablet, a coworker glancing at your screen at the office. It’s a simple, useful layer of privacy.

But don’t treat it as a safe place for data you wouldn’t otherwise put into ChatGPT. Passwords, card numbers and personal documents are still a bad idea in any chat, locked or not. Locking solves snooping, not the risk of sending sensitive data to an online service.

What to check after it reaches your account

Once the feature shows up, lock a test conversation first and make sure you can reopen it with both the PIN and biometrics. Then check that the chat no longer appears in your regular history or search. Do this before you move anything you really care about into the Locked Chats section, so you know the PIN works before you depend on it.

If you want to keep up with other ChatGPT changes, read about ads coming to the image generator and about the ChatGPT Chrome extension that reads your tabs.

See this article as a Web Story