Google published a short Gemini app post, dated November 16, 2026, saying it is giving users more control over when chat history is used to train generative AI models. New settings, the company said, let people keep their Gemini chat history while opting out of that training. Users will still be able to change how long history is saved and to review or delete history manually. "We'll start rolling out these changes globally in the Gemini app in the coming weeks," the post said.
The same post announced a second change that Google tied to safety. "Now, if you turn your new History setting off, future chats will be saved for around two months; they're used to respond to you, process your feedback, and protect Google, its users, and the public." The November 16 post does not mention a previous retention period by length. Jagmeet Singh, writing for TechCrunch on November 17, 2026, reported that the two-month window is a significant increase from a previous 72-hour retention period mentioned in Google's Gemini Apps Privacy Hub, according to an archived version of that page from November 12. That comparison is TechCrunch's, based on the archived help page, not a figure restated in the blog post.
Google's updated Gemini Apps Privacy Hub, which was also read for this package, says that even if Gemini History is turned off, future chats will still be saved for around two months and used to respond to the user, process feedback, and protect Google, its users and the public, with help from human reviewers. The hub says temporary chats are not used to train Google's AI models, and that if History is off and the user does not submit feedback, Google also does not use those future chats to train generative AI models. A related help page on managing history repeats the two-month period and notes that this retained material will not appear in the user's Gemini History.
TechCrunch wrote that until now, history and training were tied to a single "Keep Activity" setting, and that users will now be able to keep conversations in history without using them to train models. The outlet said Google told it that both new settings will be turned off by default for people who previously disabled Keep Activity, while those who had Keep Activity on will keep chat history enabled and can decide separately whether conversations are used for training. TechCrunch also wrote that Google did not answer questions about what prompted the longer retention or how conversations would be accessed or reviewed, and instead pointed to the blog post. The same article compared the two-month window with periods that OpenAI and Anthropic describe for temporary or incognito chats, saying those companies may keep such chats for up to 30 days, with possible longer retention for legal or safety reasons.
The story is in this run's two-day window because TechCrunch's November 17 report is a qualifying source report under the pipeline's rule that a later independent write-up can bring in a slightly earlier official post. The first-party blog remains the source of record for the new controls and the "around two months" retention when History is off. The 72-hour comparison, the default behavior for prior Keep Activity users, and the OpenAI and Anthropic comparisons come from TechCrunch and should be read as that outlet's reporting.