Google Gemini rolls out user-editable memory summaries

By Amara Osei

Google says Gemini’s “Memory summary” is available on the web to Google AI subscribers in US English who enable Keep Activity and Memory.

Users can open it through Settings > Personal Intelligence or at gemini.google.com/personalization-settings. Google says the tool lets them “view, edit, and delete what Gemini Apps understands about you.”

What does the summary show?

Google’s page opens with a categorized list of stored details. The source identifies five categories, each containing several bullet points with specific information:

• The Essentials: a “snapshot of your core background, location, and key relationships.”
• Career & Professional Projects
• Travel & Daily Gear
• Media & Entertainment
• Fitness & Lifestyle

Google’s description provides no total count of details, no selection criteria and no statement that the visible list exhausts every piece of information Gemini may retain.

Users can inspect the categories and individual bullets before deciding what to change.

How can users edit it?

A text field at the bottom of the list accepts the instruction “Add or remove details.” Google provides two examples:

• “Add to my memory that I speak Japanese.”
• “Remove my old phone number from your memory.”

Google also says: “Gemini may ask follow-up questions if your request is ambiguous to ensure your summary remains accurate.” The supplied explanation does not define ambiguous requests or say whether a correction can proceed without an answer.

The pencil icon in the top-right corner enables broader edits. A three-dot overflow menu contains “Memory history,” which shows “when specific information was added or modified.”

The described history provides a timing record. Google does not say whether users can compare old and new values or whether deleting a detail also removes its history entry.

Users can change individual facts and inspect when they changed, although the available description stops short of documenting a full revision trail.

Who has access now?

Google officially limits access to Google AI subscribers using US English with Keep Activity and Memory enabled. The web route is the only confirmed access point in the supplied material.

Google’s mobile rollout remains unconfirmed: the review did not find the feature on Android or iOS, although it expects a similar experience there. Google provides no launch date, app-version requirement or explanation of whether mobile eligibility will match the web.

The term “Google AI subscribers” also leaves individual plan names and prices unspecified. Web users have the only confirmed route to the feature.

What else is Google developing?

The source also points to three separate Google developments around Gemini. The provided material gives only headline-level detail:

• Google announced Gemini 4 Carbon with “significant gains across coding and agentic tasks,” meaning multi-step AI work.
• Google’s latest Gemini design comes to all Android devices, while Gemini app upgrades cover file generation.
• Google is adding more Connected Apps (service connections) to Gemini and Search with a new developer platform.

These items do not expand the stated scope of Memory summary. The provided material gives no benchmark for Gemini 4 Carbon, no rollout schedule for the Android design and no dates, markets or feature list for Connected Apps.

For marketing teams, the practical test is whether users can find and correct inaccurate Gemini details before the assistant uses them in later interactions. Google has delivered a web control, but not a mobile timetable, named subscription tiers or a documented revision trail.