ChatGPT Android Code Hints at Locked Chats Feature

A Privacy Feature Hiding in Plain Sight

A new wave of ChatGPT updates is drawing attention not for what OpenAI has officially announced, but for what has been found buried inside the app's code. According to reports, the latest version of the ChatGPT Android app contains unused strings and interface elements suggesting the company is building a way to lock individual conversations behind a PIN code or biometric authentication, such as a fingerprint scan 13. The feature appears to still be inactive, meaning it hasn't been switched on for public use, but its presence in the code strongly signals that OpenAI is testing a privacy-focused option that would let users shield sensitive chats from anyone who picks up their phone 13.

Why Locking Chats Would Matter

As ChatGPT has evolved from a novelty chatbot into a daily-use assistant, users increasingly store personal, financial, medical, and work-related information in their conversation history. A chat-locking mechanism would bring ChatGPT in line with messaging apps and note-taking tools that already offer similar protections, giving users more granular control over who can see specific threads rather than relying on a single device-level passcode. The fact that this capability is emerging specifically on Android, and tied to fingerprint or PIN verification, suggests OpenAI is treating individual conversations as data worth protecting on the same level as banking or messaging apps 13.

Part of a Broader Feature Push

This discovery comes amid a broader flurry of updates to the ChatGPT app ecosystem. OpenAI recently rolled out a feature that turns users' personal photos into custom sticker packs usable in iMessage and WhatsApp, expanding the app's role from pure assistant to a creative and social tool 2. Separately, coverage has highlighted a new capability that allows ChatGPT to draw on browser history in combination with tools like Codex, a function some commentators have likened to Microsoft's Windows Recall, but without relying on constant screenshots 4. That comparison points to growing unease about how much passive data these AI assistants are quietly accumulating in the background.

Growing Scrutiny Over Autonomy and Access

Adding to that unease, at least one report describes an AI capability that can read a user's iMessages and even reply on their behalf without explicit permission for each action, a development that left some users confused about how much control they were actually granting the assistant 5. Taken together, these stories reflect two parallel trends: ChatGPT is being woven more deeply into everyday communication and browsing habits, while simultaneously facing pressure to offer stronger safeguards for the sensitive data that deeper integration inevitably creates. Whether the locked-chats feature ships soon, and in what form, remains unconfirmed, but its appearance in the code suggests OpenAI is aware that expanded functionality needs to be matched with expanded privacy controls.