OpenAI Launches Locked Chats for ChatGPT Privacy
OpenAI introduced Locked Chats for ChatGPT, allowing users to secure sensitive conversations with a PIN or biometric authentication to hide them from history and search.
OpenAI has released a Locked Chats feature for ChatGPT, enabling users to hide sensitive conversations from regular chat history, pinned chats, and search results. The tool moves selected conversations and associated files into a protected space under the Projects menu, requiring a four-digit account PIN or biometric authentication—such as Face ID and Touch ID—for access.
The feature is rolling out across most ChatGPT tiers, including Free, Go, Plus, Pro, Business, Enterprise, and Edu plans, and is available on the latest macOS and Windows apps. However, the functionality is excluded from ChatGPT for Healthcare.
OpenAI clarified that Locked Chats provide local privacy on shared devices but do not implement end-to-end encryption. The feature does not alter data retention policies, and conversations may still be used for model training if the user has enabled the setting to improve the model for everyone. To prevent information leakage, locked chats maintain a separate memory from regular conversations.
Users are warned that resetting a forgotten account PIN results in the permanent deletion of all locked chats and associated content. This process is irreversible and cannot be recovered by OpenAI Support, though changing a known PIN preserves the data.
