ChatGPT adds PIN-protected locked chats for sensitive conversations
OpenAI, the developer of the ChatGPT AI chatbot, is rolling out a «locked chats» feature that hides sensitive conversations behind a 4-digit PIN. It is not end-to-end encryption and does not change data retention rules.

OpenAI, the company behind the ChatGPT AI chatbot, has begun rolling out a new feature called locked chats. It lets users shield private conversations from prying eyes: once enabled, people can decide whether a past exchange shows up in their regular chat history, pinned chats and conversation search. The rollout is gradual.
How it works
Accessing locked chats requires a 4-digit PIN. On an iPhone or Android phone, authentication systems such as Face ID can be used as well. Either way, the private conversations are grouped under a new project titled "Locked chats."
OpenAI presents the feature mainly as a way to keep people in public places, or someone who borrows your phone, from seeing your conversations.
Limitations
Locked chats are not end-to-end encryption, and they do not help users avoid OpenAI's legal obligations if they run into trouble with an employer or the law. The company states this explicitly on a newly published support page, which also says the feature does not automatically change model training settings or how long OpenAI retains private conversations.
Resetting the account PIN for any reason deletes all locked chats, along with any files uploaded to their project space. "This cannot be undone, and OpenAI Support cannot recover the deleted content," the company warns.
Availability
If the option is not visible right away, it may take a few days to arrive in an account. It will be available to free users as well as subscribers to the Go, Plus and Pro plans. On Android and iOS, voice mode can be used together with locked chats, but Codex conversations cannot be locked with the feature.


