Digital Transformation · 18 November 2026
Gemini chats kept two months even with history off
Google is extending how long Gemini chats are retained, from 72 hours to two months, and TechCrunch reports this applies even when chat history is turned off. The update is set to roll out globally over the coming weeks.
What happened
Google is lengthening how long it keeps conversations that users have with its Gemini assistant. Chats that were previously retained for 72 hours will now be kept for two months, and TechCrunch reports that this applies even when a user has chat history turned off.
The update is set to roll out globally over the coming weeks. This briefing draws on the TechCrunch report as its single source, so the details below are limited to what that coverage describes.
Why it matters
The most consequential detail is the interaction with the history setting. Users who switch chat history off are likely to understand that as a instruction not to keep their conversations. When retention continues regardless of that choice, the control's label and its actual behaviour no longer match, and that gap is where trust is won or lost.
For leaders deploying AI assistants inside organisations, this is a reminder that data retention is part of the user experience, not only a back-end policy. Employees and customers form expectations from what the interface promises at the moment of choice. Any AI product rollout that makes privacy claims needs those claims to be verified against how the system actually behaves.
By the numbers
- 72 hours was the previous retention period for Gemini chats.
- Two months is the new retention period, applying even with chat history off.
- 17 November 2026 is the date TechCrunch published its report on the change.
The Renascence take
Most coverage will frame this as a data-retention story, but the real issue is expectation design. A switch labelled "history off" is a promise about the experience, and a promise that quietly stops being true does more damage to trust than most feature launches can repay. The behavioural principle underneath is simple: people act on the mental model the interface gives them, not on the policy document behind it.
Treat every "off" control in your AI products as a contract with the user, and audit it against what the data pipeline really does. Write retention periods in plain language at the point where people make the choice, not in a settings page they will never open. If your product team cannot explain a setting in one sentence that matches the system's behaviour, the setting is not ready to ship.
Sources
This briefing was written by our Newsdesk, synthesising reporting from the outlets below. Follow the links for the original coverage.
FAQ
Questions we get on this topic
More in Digital Transformation
Stay ahead of CX
Get the signal, not the noise.
The stories shaping customer experience — plus the Journal and Experience Loom — in your inbox.


