OpenAI Adds Locked Chats Feature for Enhanced Privacy in ChatGPT
On Tuesday, OpenAI introduced a new feature allowing users to lock their ChatGPT conversations using a four-digit PIN or biometric authentication like Face ID and Touch ID. This enhancement is intended to protect sensitive discussions from being viewed by others who may have access to the user's device. Unlike temporary chats that disappear after use, locked chats can be revisited at any time but are stored separately in a dedicated section called 'Locked Chats.' The feature does not provide end-to-end encryption or change how long OpenAI retains chat data; it also doesn't prevent information from being remembered by the AI model before locking. Locked chats maintain their own memory, which is isolated from regular conversations to avoid leaking new information into other chats. Available across multiple ChatGPT subscription tiers but not on Healthcare plans, this feature aims to address growing concerns about privacy in personal and sensitive discussions facilitated through AI chatbots.
