OpenAI is working on a ‘Locked Chats’ feature that will hide selected ChatGPT conversations behind a PIN or biometric authentication. References to this feature were found in the code of version 1.2026.230 of the Android app, reports Android Authority.
A locked chat will not appear in the standard sidebar, history or ChatGPT search results. To access its name and content, you will need your account’s PIN or a fingerprint.
Once opened, the conversation will not automatically switch to normal mode. The chat will remain locked until the user removes the protection themselves. This should be useful on devices that may be accessed by family members or colleagues.
Locked Chats will also have separate memory handling rules. Locked conversations will be able to use information that ChatGPT has remembered from both normal and locked chats. However, new memories from locked conversations will only be used in other locked chats and will not affect normal ones.
Locking a chat will not delete any information that ChatGPT may have previously memorised from it. Therefore, the feature will create an additional barrier to viewing the chat on the device, but will not clear the service’s existing memory.
The new mode will differ from temporary chats. Temporary Chats are not saved in the history, do not create memories, and are deleted from OpenAI’s systems within 30 days. Locked Chats will remain in the account and retain the context to continue the conversation.
OpenAI has not yet officially announced Locked Chats. The presence of ready-made text descriptions within the app may indicate active development, but does not guarantee a public launch of the feature or its release in the near future.
As a reminder, ChatGPT already allows you to archive conversations, removing them from the sidebar without deleting them. However, the archive is not protected by a separate PIN code or biometrics.
Share:
Посилання скопійовано