OpenAI has introduced a new feature for ChatGPT that allows users to lock individual chat conversations to protect sensitive information. This capability is secured through a four-digit PIN or biometric authentication such as Face ID or Touch ID on compatible devices. Unlike Temporary Chats, which do not appear in chat history and are session-based, Locked Chats remain accessible to users over time but are separated from everyday conversations to enhance privacy when sharing devices with others.

The feature aims to address user needs for privacy in scenarios such as personal discussions or surprise planning, where sharing the device might otherwise expose private content. Once locked, chats move to a dedicated “Locked Chats” section, which is distinct from regular chat history, pinned chats, and search results. Messages sent after locking a chat will update a separate memory exclusive to the locked section, preventing information from leaking into other chats.

It is important to note that locking a chat does not offer end-to-end encryption or alter how OpenAI stores or handles logged conversations. The company highlights that this feature does not change data retention policies, how the model learns from chats, or the “Improve the model for everyone” settings. Additionally, resetting the PIN deletes all locked content, and certain chat types like group chats, temporary chats, and those with scheduled tasks cannot be locked.

Locked Chats will be available across a wide range of ChatGPT plans including Free, Go, Plus, Pro, Business, Enterprise, and Edu, but it will not be supported on ChatGPT for Healthcare. The rollout began on October 20, 2026, though the feature may not yet be live for all users. This update adds a useful privacy layer as ChatGPT continues to be used for more intimate and personal interactions.