OpenAI adds PIN and Face ID locks to individual ChatGPT conversations

OpenAI has begun rolling out a feature that lets users lock individual ChatGPT conversations behind a four-digit PIN or, on supported devices, Face ID or Touch ID. According to TechCrunch, the company announced the capability on Tuesday and framed it as a way to keep sensitive exchanges private when someone else has access to your screen or device.

How locking works

Once a conversation is locked, it moves into a dedicated "Locked Chats" section that sits apart from the regular chat history, pinned conversations and, notably, chat search results. Anyone scrolling through the account or running a search won't stumble across it.

That makes the feature different from Temporary Chats, which create a separate session that never enters the history at all. Locked chats persist — users can return to them at any time. TechCrunch describes the intended scenario as handing your device to a friend or family member to show them something in another chat without exposing a personal conversation or, in OpenAI's own example, the planning for a surprise.

A separate memory space

With Memory enabled, the Locked Chats section gets its own memory, shared among conversations inside that section but walled off from the regular Memory feature. Messages sent after a chat is locked can update the locked memory, but never the main one, so facts picked up in a protected conversation won't surface later in ordinary chats.

One caveat: information ChatGPT remembered before a chat was locked stays in regular memory, so the wall only applies going forward.

What it doesn't do

OpenAI is unusually explicit about the limits. Per TechCrunch, locking a chat does not create end-to-end encryption or a separate encrypted vault, does not change how long the conversation is retained, does not remove previously remembered information, and does not alter the "Improve the model for everyone" setting. In other words, this is an access control on the interface, not a change to how the underlying data is stored or handled.

There is also a sharp edge for forgetful users: resetting the PIN permanently deletes every locked chat, file and other piece of content protected by the previous PIN. No recovery path is described.

Availability and restrictions

The feature can't be applied everywhere within the product. Temporary chats, archived chats, group chats, conversations created from shared links, chats with scheduled tasks, and chats containing meeting recordings are all excluded.

Locked Chats is reaching users on ChatGPT Free, Go, Plus, Pro, Business, Enterprise and Edu plans, but is not available on ChatGPT for Healthcare. Because the rollout began Tuesday, it may not be visible to every account yet.

Why it matters

ChatGPT has become a place where people discuss health, relationships, finances and work, frequently on devices they share or hand over. Until now the app's privacy options were essentially binary: delete the conversation or leave it in plain sight of anyone holding the phone. A per-conversation lock closes a real usability gap and acknowledges how personal these assistants have become — especially as memory features accumulate intimate context over time.

It's equally important to be clear about what this isn't. As TechCrunch points out, the feature does nothing to address concerns about how sensitive information is stored or processed on OpenAI's side. Users who trust the company with their data but not with shoulder-surfing family members get genuine protection; users worried about retention periods or training-data usage get nothing new. As AI assistants absorb more of daily life, the gap between interface privacy and data privacy is likely to become a more prominent part of the debate — and this launch is a good illustration of exactly where that line sits today.