ChatGPT now lets you lock your chats

Introduction

OpenAI has rolled out a new privacy tool for ChatGPT users: Locked Chats. By securing selected conversations with a four‑digit PIN, Face ID, or Touch ID, the feature aims to keep personal or sensitive discussions hidden from anyone who might glance at your device. While it adds a convenient layer of on‑device protection, it does not change how OpenAI stores or processes the underlying data.

How Locked Chats Work

When you lock a conversation, it moves to a dedicated Locked Chats section, separate from your regular chat history, pinned chats, and search results. The lock can be applied to any standard chat (except temporary, archived, group, shared‑link, scheduled‑task, or meeting‑recording chats). Once locked, the chat can be reopened at any time by entering the PIN or using biometric authentication.

Key characteristics

• Access control – A four‑digit PIN, Face ID, or Touch ID is required to view the chat.
• Separate memory – Locked Chats maintain their own memory store, isolated from the main Memory feature. New messages added after locking update this isolated memory but do not affect the regular memory pool.
• Deletion – Resetting the PIN permanently erases all locked content tied to the previous PIN.
• Availability – The feature is being rolled out across all ChatGPT plans (Free, Go, Plus, Pro, Business, Enterprise, Edu) but is not yet supported in the Healthcare version.

What Locked Chats Don’t Do

It’s important to understand the limits of the new feature:

• No end‑to‑end encryption – Chats remain stored on OpenAI’s servers under the same security model as other conversations.
• No change to retention – Locked chats are retained according to the same policies as regular chats; they are not placed in a special vault.
• Model training – Content in locked chats can still be used to improve the model unless you have opted out of data sharing globally.
• No impact on other features – Locked chats are unavailable for temporary sessions, group chats, or chats generated from shared links.

Practical Tips for Users

1. Use locks for truly sensitive topics – Treat Locked Chats as a screen‑privacy measure rather than a guarantee of data confidentiality.
2. Combine with global privacy settings – If you wish to keep any data from training, disable the “Improve the model for everyone” option in your account settings.
3. Regularly review locked content – Since retention policies remain unchanged, consider deleting old locked chats you no longer need.
4. Leverage biometrics where possible – Face ID or Touch ID offers a smoother experience than remembering a PIN.

Conclusion

Locked Chats give ChatGPT users a straightforward way to hide specific conversations from prying eyes on their devices, enhancing on‑screen privacy without altering the backend handling of data. While the feature is a welcome addition for personal use cases—such as planning surprises or discussing confidential matters—it does not replace robust encryption or data‑governance measures. Users should continue to manage their broader privacy settings and stay aware of what information OpenAI retains and how it may be used.