OpenAI Adds Locked Chats to ChatGPT, Giving Users More Control Over Private Conversations

The new Locked Chats feature provides a mechanism for users to separate selected conversations from their main chat history. This is particularly useful for shared devices, where users might discuss private matters or sensitive information and want to restrict casual viewing by others who use the same phone or computer.

Unlike Temporary Chats, which do not appear in history, Locked Chats remain accessible to the account holder after authentication. This distinction offers a way to retain conversation content while adding a layer of protection against unauthorized access within the application. Once locked, conversations move to a dedicated section, separate from regular chat history and search results.

OpenAI has also implemented a separate memory arrangement for Locked Chats when the Memory feature is enabled. This allows information within locked conversations to inform other chats within that protected section without influencing the user's broader, regular Memory. This separation helps keep certain topics isolated from general ChatGPT interactions.

However, the feature's limitations are notable. Locking a conversation does not introduce end-to-end encryption, create a separate encrypted vault, or alter ChatGPT's data-retention practices. It also does not change how the model processes or remembers information, or modify the user's "Improve the model for everyone" setting. The protection primarily controls in-app access on a shared device rather than establishing a fundamentally different security environment for the data itself.