Google Extends Gemini Chat Retention To Two Months, Even With History Off
The company is extending temporary chat retention from 72 hours while introducing separate controls for conversation history and AI model training.

Google is extending how long it retains Gemini conversations even when users turn off chat history, increasing the period from 72 hours to around two months as the company introduces new privacy controls for its artificial intelligence assistant.
The change means that disabling Gemini's History setting will no longer limit the retention of future conversations to approximately three days. Temporary Chats will also be kept for around two months, according to Google's updated privacy documentation.
Google announced the changes Monday alongside a new system that allows users to keep their conversation history while separately choosing whether their chat content can be used to train generative AI models.
The company said the updates would begin rolling out globally over the coming weeks.
In its official announcement, Google said the longer retention period was intended to strengthen safety and security protections as AI systems become increasingly capable.
Under the new policy, future conversations will be saved for around two months even when users have disabled history.
Google says the retained information may be used to respond to requests, process feedback and protect the company, its users and the public.
The change represents a substantial increase from the previous 72-hour retention period.
The Register reported Tuesday that archived versions of Google's privacy documentation confirmed the earlier three-day limit.
The publication also noted that Google's announcement emphasized the new privacy controls without highlighting how much longer conversations would be stored when history was disabled.
The distinction matters because turning off chat history does not mean Google immediately deletes the information exchanged with Gemini.
Under the updated policy, conversations may remain associated with a user's Google Account during the temporary retention period, even though they do not appear in the user's visible Gemini History.
Google also says it may use human reviewers to examine certain conversations for safety and security purposes.
However, the longer retention period does not automatically mean that every conversation will be used to train Google's AI models.
One of the main changes announced this week is the separation of chat history from permission to use conversations for generative AI training.
Previously, retaining Gemini conversation history was more closely tied to allowing Google to use those interactions to improve its AI systems.
The new settings allow users to preserve their conversations for future reference while opting out of generative AI model training.
That gives people who want to revisit previous conversations greater flexibility without requiring them to authorize that particular use of their data.
Google's documentation states that future chat content will not be used to train generative AI models when the training setting is disabled, except in certain circumstances, including when users voluntarily submit feedback.
Temporary Chats are also excluded from generative AI model training, although Google still retains them temporarily for operational and safety purposes.
There is another important exception involving human review.
According to Google's Privacy Hub, conversations reviewed by human reviewers and associated information can be retained for up to three years, even if users subsequently delete their history.
Google says conversations sent to service providers for review are disconnected from users' accounts.
Submitting feedback about a Gemini response can also result in the associated conversation, attachments and other relevant information being reviewed and retained for up to three years.
That information may be used to maintain, improve and develop Google services, including machine-learning technologies.
The new settings therefore distinguish between several different activities: saving conversations in a user's visible history, retaining information for operational or security purposes, and using chat content to train generative AI models.
Switching off one of those functions does not necessarily disable the others.
Users who want to review their choices can open Gemini, go to Settings and select History.
Google also provides access to conversation settings through Gemini History, where users can review stored activity, delete conversations and manage retention preferences.
With the new controls, users can choose whether their chat content, including text and uploaded files, may be used for generative AI training.
Google's help documentation says the new training controls are off by default, although individual users may see different options while the global rollout is underway.
For users who leave chat history enabled, Google says the default automatic deletion period remains 18 months. They can change that period to three months or 36 months, or choose not to automatically delete their history.
However, those settings do not override every separate retention requirement.
Google also warns that deleting Gemini history does not necessarily remove information saved through connected Google services, which may follow their own privacy and retention policies.
The company advises users not to enter confidential information they would not want a reviewer to see.
© Copyright IBTimes 2026. All rights reserved.

