What the labels tell you
A Temporary Chat is not a blanket promise that nothing is retained. OpenAI says a temporary chat “stays out of your chat history” while it remains temporary, but it “may keep a copy” for up to 30 days for safety purposes. If the chat is Personalized, it “can use” existing memories, custom instructions, and plugins.
These labels answer narrower questions than a general claim that a conversation is completely private.
What each label means
| Label | What OpenAI’s wording indicates | What the label alone does not establish |
|---|---|---|
| Temporary Chat | The chat stays out of chat history while it remains temporary. | That no copy is retained. OpenAI says a copy may be kept for up to 30 days for safety purposes. |
| Personalized for a temporary chat | The chat can use existing memories, custom instructions, and plugins. | That every listed feature was used in a particular exchange, or that temporary status prevents personalization from being used. |
“Up to 30 days” describes a possible safety-retention period, not a promised retention duration for every temporary chat. Likewise, “can use” describes a capability rather than confirming which specific memories, instructions, or plugins affected a response.
How to check the relevant settings
First, confirm that the interface identifies the conversation as temporary. Then check personalization separately: temporary history status does not, by itself, answer whether existing memories, custom instructions, or plugins may be used.
Keep three questions distinct:
- Chat history: OpenAI says a temporary chat stays out of history while it remains temporary.
- Safety retention: OpenAI says a copy may be kept for up to 30 days.
- Personalization: A personalized temporary chat can use existing memories, custom instructions, and plugins.
OpenAI’s quoted descriptions do not provide a menu path, so check the controls actually presented in the interface rather than assuming a particular navigation route.
What you must still confirm yourself
The Temporary Chat label does not mean “never retained,” and it should not be treated as proof that personalization is disabled. The Personalized label indicates possible use of saved context, but it does not identify what was used in a specific conversation.
Before relying on either label, confirm the active settings in your interface and review OpenAI’s current Help Center guidance. Do not assume that a differently named label, such as “Private,” has the same history, retention, and personalization behavior unless its documentation explicitly defines those points.