Home · Technology · Nov 10 archive

Google Gemini Launches 'Memory Summary' to Show User Data

Confirmed

Technology Desk

In Short: The launch of 'Memory summary' addresses a common grievance about Gemini's memory, providing a clearer picture of what the chatbot remembers about users.

Google; Google Gemini logos
Logos: Google (Public domain); Google Gemini (Public domain)

How to Use Google Gemini Step by Step (It Can Do Your Work Now). Teacher's Tech

Google has introduced a new feature called 'Memory summary' for its Gemini AI assistant, allowing users to see, edit, and delete the information the AI has collected about them.

The 'Memory summary' feature is accessible via Settings > Personal Intelligence or directly at gemini.google.com/personalization-settings.

The text field at the bottom of the summary allows users to add or remove details. For instance, Google says Gemini may ask follow-up questions if a request is ambiguous to ensure the summary remains accurate.

The feature also includes a 'Memory history' view that shows when specific information was added or modified.

Users can ask Gemini directly in plain language to correct or remove details, such as telling it to drop an old phone number.

While the feature is officially limited to Google AI subscribers using Gemini in US English, it was found working on Gemini’s desktop website with a free account outside the US, without using a VPN.

However, the feature was not available in the Gemini mobile app, even with the latest version installed.

Google designed the 'Memory summary' feature to help users review, correct, and remove details Gemini uses for personalized responses.

The summary sorts what Gemini knows into categories, giving users greater visibility and control over personal information gathered through previous conversations.

Users can also turn Memory off through the Personal Intelligence settings if they no longer want Gemini to learn about them.

What this adds

The 'Memory summary' is currently limited to Google AI subscribers using Gemini in US English, as reported by AndroidPure.

The feature was unexpectedly accessible on the Gemini desktop website with a free account outside the US, but not on the mobile app.

Background

Google's Gemini 4 Carbon model excels in cybersecurity and coding benchmarks, outperforming competitors like Claude and GPT-6 Astra.

Gemini, sometimes marketed as Google Gemini and originally known as Bard, is a generative artificial intelligence chatbot and virtual assistant developed by Google. It is powered by the family of large language models (LLMs) of the same name, previously based on LaMDA and PaLM 2.

What's confirmed

What's still developing

Sources