Enhanced Security for Private Conversations
OpenAI appears to be working on a significant privacy upgrade for its popular AI assistant. Recent findings in the latest Android version of ChatGPT suggest that users may soon be able to secure specific, sensitive conversations using PIN codes or biometric authentication like fingerprint scans.
How the Feature Will Work
The potential update was uncovered by researchers who spotted inactive code references within the application. These snippets suggest that users will have the ability to explicitly "lock" certain threads. According to the discovered strings, this functionality is designed to:
- Hide chat titles from the main sidebar.
- Remove content from standard history views.
- Exclude locked conversations from regular search results.
This approach differs from the existing "temporary chat" mode. While temporary sessions are erased immediately upon closing, "locked chats" seem intended for more permanent storage while maintaining a higher degree of confidentiality.
Privacy Beyond Device Security
While most modern devices already offer system-level protection through facial recognition or screen locks, this feature adds an extra layer of defense specifically within the app. It provides a safeguard for users handling delicate information, such as:
«From embarrassing medical questions to queries about surprise gifts for family members to drafts of resignation letters for your boss, there are probably quite a few AI chats you want to make sure no one else ever sees.»
At this stage, questions remain regarding how these secured interactions will integrate with ChatGPT’s existing "memory" feature. It is currently unclear if the AI will retain context from these locked sessions for use in other chats. OpenAI has not provided an official release date for this functionality, and as with all pre-release code, there remains a possibility that the project could be modified or shelved before a public launch.