ChatGPT now lets you lock private chats with a PIN—but forgetting it could erase them for good

OpenAI is giving ChatGPT users a way to tuck sensitive conversations out of sight—but the new feature comes with a steep penalty if you forget your PIN.

The company began rolling out locked chats on Tuesday, which let users keep selected conversations in their account while hiding them from their regular chat history, pinned chats, and conversation search. OpenAI says the feature is designed to keep sensitive conversations out of view when someone shares their screen or uses a shared device.

Locked conversations move to a dedicated area under Projects > Locked chats, and opening them requires a four-digit account PIN. On supported mobile devices, users can also unlock them with Face ID, Touch ID, or their phone’s device unlock method.

Users should think carefully before resetting a forgotten PIN. Doing so permanently deletes all locked chats, any memories stored only within locked chats, and files or other content protected only by the previous PIN. According to OpenAI, the deletion cannot be undone, and its support team cannot recover the content. Workspace administrators also can’t reset a user’s PIN or unlock their chats.

Changing a PIN you still remember, by contrast, doesn’t delete anything.

What locked chats don’t do

Locking a chat changes where a conversation appears and how it’s opened—not how OpenAI handles it. According to the company, the feature doesn’t create end-to-end encryption or a separate encrypted vault, doesn’t change how long OpenAI retains the chat, and doesn’t alter a user’s “Improve the model for everyone” setting. It also doesn’t remove anything ChatGPT remembered before the conversation was locked, and it can’t override workspace policies, safety requirements, or legal and compliance obligations.

In ChatGPT Enterprise and Edu workspaces, locked chats are still included in the conversation logs available through OpenAI’s Compliance Platform.

Locked chats do get some separation when it comes to memory. When Memory is turned on, information from new messages sent after a chat is locked can be remembered for use in other locked chats, but it doesn’t carry over into regular conversations.

Where locked chats are available

The feature is rolling out gradually across ChatGPT Free, Go, Plus, Pro, Business, Enterprise, and Edu plans. On desktop, users can lock eligible conversations in the new ChatGPT app for macOS or Windows, in either Chat or Work mode. Voice works in locked chats on the web and in the iOS and Android apps where it’s available, but not in the desktop app.

Not every conversation qualifies. Temporary chats, archived chats, group chats, chats created from shared links, and chats with scheduled tasks or meeting recordings can’t be locked, among others. Codex conversations and local Work conversations are also excluded, and the feature isn’t available on ChatGPT for Healthcare.

Locking a chat also doesn’t delete an existing shared link. Anyone with that link can still view the previously shared content, though new messages sent while the chat is locked won’t be added. To revoke access, users need to delete the link separately. Shared links have been a privacy sore spot before: last year, OpenAI ended an option that let shared ChatGPT conversations be discovered on Google after user complaints.

Files uploaded or created within a locked chat are protected along with the conversation and hidden from Space, ChatGPT’s area for saved files, until the chat is unlocked. Files that were already saved in Space before being added to the chat, however, remain available there.

Locked chats vs. Temporary Chat

Locked chats keep sensitive conversations out of view, but they don’t stop OpenAI from processing them under its existing policies. Users who want to manage how their conversations are used to improve the model need to review their data controls separately.

Temporary Chat offers a different set of protections. According to OpenAI, temporary conversations don’t appear in chat history, don’t create or update memories, and aren’t used to improve its models while they remain temporary—though the company may keep a copy for up to 30 days for safety purposes. Locked chats, by contrast, remain saved and follow a user’s existing model-improvement setting.

A bigger privacy question for AI chatbots

The distinction matters as AI chatbots become part of daily life and questions about how companies handle personal information continue to grow. OpenAI said at its DevDay event last month that ChatGPT now has 1.2 billion weekly users. The chatbot’s reach into personal data is expanding, too: In August, Fortune reported that ChatGPT can now search a Mac user’s Apple Messages once granted access. And the issue isn’t unique to OpenAI. In July, a trove of users’ seemingly private conversations with Anthropic’s Claude showed up in Google search results.

A Stanford study published in October 2025 examined the privacy policies of six leading U.S. developers: Amazon, Anthropic, Google, Meta, Microsoft, and OpenAI. The researchers found that all six appeared to use users’ chat data to train and improve their models by default, although practices and opt-out options varied.

The study found that developers may collect and train on sensitive personal information users share with chatbots, including information contained in uploaded files. The researchers also flagged concerns about lengthy data-retention periods, the handling of children’s data, and a lack of transparency in companies’ privacy policies.

“As a society, we need to weigh whether the potential gains in AI capabilities from training on chat data are worth the considerable loss of consumer privacy. And we need to promote innovation in privacy-preserving AI, so that user privacy isn’t an afterthought,” Jennifer King, a privacy and data policy fellow at the Stanford Institute for Human-Centered AI and the study’s lead author, said in a summary of the findings published by Stanford.

This story was originally featured on Fortune.com