OpenAI is upgrading ChatGPT's temporary chat feature, allowing users to maintain personalization while keeping conversations separate from their main account. In this mode, users start fresh, with no past chats or memories influencing the interaction, even if memory is usually active. The temporary chat will still adhere to any Custom Instructions set by the user and won’t appear in the chat history or contribute to model improvements. With the update, temporary chats can now access user preferences for memory, chat history, style, and tone, all while retaining their temporary status. Users can disable this personalization if they choose. For safety, OpenAI may keep a record of these chats for up to 30 days. To initiate a temporary chat, users simply select the "Temporary" button at the top right of the chat window. In addition to this, OpenAI has introduced an age prediction model that evaluates user behavior and chat patterns to estimate age. The model isn't foolproof—some adult users may be misidentified as teenagers, leading to account restrictions. Restricted users can still interact, but they face limitations on topics like violence or viral challenges. Adults wrongly categorized can verify their age to regain full access to ChatGPT's features.