Stay up to date with notifications from The Independent

Notifications can be managed in browser preferences.

Google’s Gemini AI now automatically ‘remembers’ your past conversations

AI chatbot now recalls a user’s favourite hobbies, creative projects and personal preferences

Anthony Cuthbertson
Thursday 14 August 2025 05:50 EDT
Comments
Google's Gemini can now recall interactions with users without being prompted
Google's Gemini can now recall interactions with users without being prompted (Getty Images)

Google has introduced an update to its AI chatbot Gemini that allows it to remember past conversations and personal details of users.

The new feature, which is automatically enabled, is aimed at making conversations feel more personalised.

The artificial intelligence will now recall a user’s favourite hobbies, creative projects or personal preferences – similar to some of its rivals.

This new ‘Personalised Context’ setting is on by default, however users are able to manually switch it off through the Gemini app.

“We’re introducing a new setting that allows Gemini to learn from your past conversations over time,” Michael Siliski, a senior director at Gemini, wrote in a blog post announcing the new update.

“When this setting is on, Gemini remembers key details and preferences you've shared, leading to more natural and relevant conversations, as if you're collaborating with a partner who's already up to speed.”

Another new feature is aimed at more privacy-conscious users, allowing single-use conversations that do not store any information.

Similar to Google’s incognito mode in the Chrome web browser, the conversations in ‘Temporary Chats’ do not show up in a user’s chat history and are not saved to Gemini Apps Activity.

Any interactions with Gemini that are made through Temporary Chats, which can be enabled via a new button next to ‘New Chat’ in the side menu, will only be saved for up to 72 hours in order to process optional feedback.

Both Temporary Chats and Personalised Context are available though an update that is rolling out to users of Gemini 2.5 Pro in selected countries before becoming more broadly available.

A similar feature was introduced to OpenAI’s ChatGPT in April, which some have linked to an increase in “delusional” chats from the AI.

OpenAI has said it is working with experts to introduced more safety measures and add additional mental health guardrails for vulnerable users.

Join our commenting forum

Join thought-provoking conversations, follow other Independent readers and see their replies

Comments

Thank you for registering

Please refresh the page or navigate to another page on the site to be automatically logged inPlease refresh your browser to be logged in