Google is updating its Gemini AI chatbot with a new feature that allows it to recall past conversations without being prompted. When this setting is turned on, Gemini will automatically retrieve key details and preferences from previous chats to personalize its output.
This update builds upon the “remember” feature introduced last year, which lets users ask Gemini to recall their personal preferences and interests. Now, Gemini won’t need prompting to use this information. For instance, if a user has previously discussed ideas for a YouTube channel on Japanese culture, Gemini may suggest creating content about trying Japanese food in future conversations.
The update aims to improve safety by providing users with more control over their conversation history. Google is working to detect mental or emotional distress and respond accordingly. However, the company acknowledges concerns about “delusional” chats and is constantly improving its safeguards.
To turn on this feature, users can disable it through the Gemini app settings by selecting Personal Context and toggling off the “Your past chats with Gemini” option. The update will initially roll out to Gemini 2.5 Pro models in select countries before expanding to more locations and other models.
Additionally, Google is making changes to its privacy settings. Starting in September, the company will rename its “Gemini Apps Activity” setting to “Keep Activity.” This setting allows Google to use a sample of user file and photo uploads to improve services for everyone. However, users can also preserve their privacy through a new “temporary chats” option, which won’t appear in recent chats or Keep Activity settings and won’t be used for training AI models.
Source: https://www.theverge.com/news/758624/google-gemini-ai-automatic-memory-privacy-update