ChatGPT Locked Chats Feature Lets You Hide Sensitive Conversations With PIN or Biometric Authentication

OpenAI Introduces Locked Chats to Protect Sensitive Conversations

OpenAI rolled out a new privacy feature for ChatGPT designed to hide sensitive conversations from prying eyes when users share devices or work in public spaces. The Locked Chats feature moves selected conversations out of regular chat history, pinned chats, and conversation search into a protected section accessible only through a four-digit PIN or biometric authentication on supported mobile devices.1

1

Users can find these protected conversations under Projects > Locked chats in the ChatGPT sidebar, with titles remaining hidden until authenticated.4

4

The feature addresses common scenarios where users might hand their device to friends or family, or use ChatGPT in crowded spaces like buses or subway platforms where others could see their screen.1

1

To lock a chat, users tap the three-dot menu in any ChatGPT conversation and select "Lock chat," then create or enter their account PIN.5

5

On iOS and Android devices, Face ID, Touch ID, or device unlock methods provide additional authentication options.4

4

Critical Limitations and PIN Reset Consequences

While Locked Chats adds a layer of privacy, forgetting your PIN carries severe consequences. Resetting a forgotten account PIN permanently deletes all locked chats, any memories stored within those conversations, and files or other content protected only by the previous PIN.3

3

OpenAI states this deletion cannot be undone, and its support team cannot recover the content—workspace administrators also lack the ability to reset a user's PIN or unlock their chats.3

3

However, users who remember their current PIN can change it without losing any locked content.5

5

The feature doesn't provide end-to-end encryption or create a separate encrypted vault.3

3

OpenAI can still access locked chats and use them for AI model training if users have opted into the "Improve the model for everyone" setting.1

1

The feature also doesn't change how long OpenAI retains conversations or override applicable workspace policies, safety requirements, or legal and compliance obligations.3

3

In ChatGPT Enterprise and Edu workspaces, locked chats remain included in conversation logs available through OpenAI's Compliance Platform.3

3

Memory Segmentation and Information Leakage Concerns

OpenAI implemented memory segmentation between regular and locked chats to prevent data privacy leaks. When the memory feature is enabled, locked chats maintain their own separate memory shared only between conversations inside the protected space.1

1

New messages sent after locking a chat can update that isolated memory but don't update regular Memory, preventing information from locked chats from appearing in standard conversations.4

4

However, the feature won't remove information ChatGPT remembered before the conversation was locked.1

1

If users discuss sensitive details before activating the lock, those details remain part of regular ChatGPT memory and can influence future responses.2

2

Users must manually navigate to the memory section and remove such information themselves.2

2

For maximum privacy, users should lock conversations before beginning sensitive discussions.

Shared Links and Temporary Access Vulnerabilities

Locking a chat doesn't delete existing shared links—anyone with a previously created link can still view the shared content, though new messages sent while the chat is locked won't be added.3

3

To revoke access completely, users need to delete the shared link separately.2

2

This represents a significant gap in the feature's protection, as information already shared remains accessible regardless of subsequent locked status.

Additional vulnerabilities exist in how the feature handles temporary access. A locked conversation can remain open in the same browser tab immediately after being locked, and previously opened devices might temporarily display its title until they refresh.2

2

Unlocking the protected project provides temporary access, but re-locking doesn't happen automatically unless users manually select Lock again when finished.2

2

Users cannot assume navigating away automatically secures the conversation again.

Availability and Conversation Type Restrictions

Locked Chats is rolling out gradually across ChatGPT Free, Go, Plus, Pro, Business, Enterprise, and Edu plans.1

1

The feature works on desktop through the new ChatGPT app for macOS or Windows in both Chat or Work mode, and voice functionality operates in locked chats on web and mobile apps where available—though not in the desktop app.3

3

Not every conversation qualifies for locking. Temporary chats, archived chats, group chats, chats created from shared links, and chats with scheduled tasks or meeting recordings cannot be locked.3

3

Codex conversations and local Work conversations are also excluded, and the feature isn't available on ChatGPT for Healthcare.3

3

Files uploaded or created within a locked chat are protected alongside the conversation and hidden from Space until unlocked, though files already saved in Space before being added to the chat remain available there.3

3

Locked Chats Versus Temporary Chat for Data Privacy

For users prioritizing comprehensive data privacy, Temporary Chat offers different protections than Locked Chats. Temporary conversations don't appear in chat history, don't create or update memories, and aren't used to improve OpenAI's models while they remain temporary—though the company may retain a copy for up to 30 days for safety purposes.3

3

Locked chats, by contrast, remain saved and follow a user's existing model-improvement setting.3

3

This distinction matters as AI chatbots integrate deeper into daily life and questions about how companies handle personal information continue growing. OpenAI announced at its DevDay event that ChatGPT now has 1.2 billion weekly users.3

3

The chatbot's reach into personal data is expanding—in August, ChatGPT gained the ability to search a Mac user's Apple Messages once granted access.3

3

Privacy concerns aren't unique to OpenAI; in July, users' seemingly private conversations with Anthropic's Claude appeared in Google search results.3

3

Locked Chats addresses visibility concerns but doesn't fundamentally change how OpenAI processes user data under existing policies.