Imagine a digital space where you can converse freely, knowing every word vanishes without a trace. That’s the promise of ChatGPT’s “Temporary Chat” – a feature designed for private, ephemeral discussions. It’s akin to a digital incognito mode, offering a clean slate for each interaction.
But what if you wanted the benefits of a personalized experience – ChatGPT remembering your preferred style and tone – *without* the conversation being saved? OpenAI appears to be quietly testing exactly that. A subtle update is emerging, hinting at a more nuanced approach to privacy.
The discovery came from AI engineer Tibor Blaho, who unearthed a hidden setting within the chatbot. This option allows users to leverage their existing customizations – the AI’s understanding of your preferences – even within a temporary, unsaved chat session.
Essentially, you could enjoy a conversation tailored to your liking, while still maintaining complete confidentiality. The setting isn’t universally available yet, remaining an experimental feature for select users. It also appears to be optional, allowing you to choose between a truly blank slate and a personalized, temporary experience.
However, a crucial detail remains: even with Temporary Chat, OpenAI may retain copies of conversations for up to 30 days for security purposes. This underscores the complex balance between user privacy and responsible AI development, a consideration that continues to shape the evolution of these powerful tools.
