[DigitalToday reporter Seung-ah Yoo (유승아)] OpenAI's ChatGPT may add a feature to lock sensitive conversations with a PIN or fingerprint authentication.

Tech outlet TechRadar reported on Aug. 23 that phrases related to "lock chat" and "protect chat" were found in inactive code in the latest Android version of ChatGPT.

The code spotted this time is not yet at a stage where it actually works, but the direction of the feature is relatively clear. The code includes wording to the effect of telling users to lock a conversation to keep the chat title and content from appearing in the regular sidebar, history and search results. It is read as a design intended to hide specific conversations separately and require additional authentication to view them.

If introduced, users would be able to add extra protection within ChatGPT separate from device locks. The key point is an added layer of security for sensitive conversations even if a smartphone or PC is already protected by a PIN, fingerprint or facial recognition. It appears to target demand for conversations users do not want exposed to others, such as personal medical questions, preparing a surprise gift for family, or a draft resignation letter to send to a boss.

ChatGPT currently has a "temporary chat" feature. It is an option for personal or sensitive conversations that is forgotten as soon as the window is closed and disappears from the account. The locked chat spotted this time, unlike temporary chat, is likely to be stored in a more persistent way.

Details of how it would work still remain. In particular, it is unclear how it would link with ChatGPT's memory feature. If memory is on, it has not been confirmed whether content entered in a locked chat would be referenced only in other locked chats, or reflected in regular chats as well. This will be a variable that divides users' perceived level of security until an actual release.

It is not yet known when OpenAI would unveil the feature. A trace in app code does not necessarily lead to an official release, and OpenAI could change course and not roll it out. Still, as ChatGPT is being used more deeply for personal work and private counseling beyond being a simple question-and-answer tool, conversation-level security features could be tied to future service competitiveness.

The code find is read as a sign that ChatGPT is considering conversation-by-conversation access controls beyond account-wide security. Whether it is actually applied, and how the memory feature, record management and search exposure scope are organized, will be points to watch.