Gemini now stores and surfaces work chats across Google Workspace

Google has updated the Gemini app to automatically store users' work conversations and surface them inside Google Workspace applications. The change means Gemini can now recall previous chats, reference documents you worked on earlier, and use that context when answering new questions inside tools like Gmail, Docs, and Meet. For anyone using Workspace heavily for day-to-day work, the practical effect is that Gemini starts to behave more like a colleague who was in the room for every previous conversation rather than a fresh session that forgets everything the moment you close the tab.

The update is significant in scale. Google Workspace has more than three billion users across its free and paid tiers, and the Gemini integration is available to all paid Workspace subscribers. That puts persistent AI memory in front of a very large number of enterprise and professional users at once.

What the feature actually does inside Workspace

When a user asks Gemini a question inside Google Docs, for example, Gemini can now pull context from previous Gemini chat sessions to inform its response. If you spent time last week asking Gemini to help draft a project proposal, and this week you open a new document and ask Gemini to help write a follow-up email about that same project, it can connect those two interactions without you needing to re-explain the background.

The memory layer also extends to documents themselves. Gemini can reference files you have recently opened or edited in Google Drive when generating responses, which means it can ground its answers in your actual work rather than generic knowledge. This is the part of the update that makes the feature genuinely useful rather than just interesting. An AI assistant that knows what documents you actually work with daily produces different outputs than one operating without that context.

Privacy settings users need to check now

The rollout has drawn attention from privacy advocates, and their concern is straightforward. Memory features that store conversation history create a new category of data that sits somewhere in Google's infrastructure. By default, Gemini chat history is stored and used to personalize responses. Users who want to limit this can turn off Gemini Apps Activity in their Google Account settings, which stops conversations from being saved. Workspace admins can also control whether the memory feature is available to users in their organization at all.

The default-on behavior is the part worth paying attention to. Many users will not know the feature has been turned on until they notice Gemini referencing a previous conversation. That is a different experience from opting into a memory feature, and the distinction matters for people handling sensitive client information or confidential internal discussions inside Workspace.

How this fits into the competition with Microsoft Copilot

Microsoft has been rolling out Copilot memory features inside Microsoft 365 over the past year, and the competitive pressure between the two platforms has been a direct driver of this update's timing. Both Google and Microsoft are making the same bet: that enterprise customers will pay a premium for AI assistants that accumulate context over time rather than starting fresh every session.

The business model behind this is worth understanding. Gemini Business and Gemini Enterprise plans, which sit on top of standard Workspace subscriptions, are priced at $20 and $30 per user per month respectively. Memory and cross-app context features are part of what Google is using to justify those prices against the standard Workspace tiers that do not include full Gemini capabilities.

What enterprise IT teams are watching

For IT administrators managing Workspace deployments at larger organizations, the memory update introduces a governance question that did not exist before. When Gemini stores conversation history, that history may contain proprietary information, client data, or internal strategy discussions. Under Google's current Workspace terms, customer data in paid tiers is not used to train Google's AI models without explicit opt-in. But stored conversation history still needs to be managed under the organization's own data retention and compliance policies.

Google has said that Workspace admins have full control over Gemini memory settings at the organization level through the Admin Console. The company's next scheduled Workspace feature update briefing for enterprise customers is expected in April 2026, where Google is likely to provide more detailed documentation on how conversation history interacts with existing Vault and data retention configurations.