LOCKE FOR GEMINI
Use Gemini across your Google work without exposing what's in it.
Gemini is different from a standalone chatbot because of where it lives: inside Google Workspace, with reach into Gmail, Docs, Sheets, and Drive. The same integration that makes it helpful — "summarise this thread," "draft a reply," "pull the figures from this sheet" — means the data flowing through it is the working material of your business, not a one-off question typed into a blank box.
That blurs the line users rely on. With a chatbot you decide what to paste; with a workspace assistant, the assistant reaches into content you didn't consciously hand it. A thread you ask Gemini to summarise might contain a client's medical detail or a colleague's compensation, and you never made an explicit decision to send it anywhere.
Locke keeps the decision where it belongs — on your device. Sensitive values are detected and masked before any prompt or selected content is transmitted, so you can lean on Gemini across your workspace without quietly routing your most confidential material to the model.
The risk with Gemini
Because Gemini is wired into Gmail, Docs, and Drive, the content it processes is already your most sensitive — and a "summarise this" action can pull regulated data into a prompt without the user ever consciously choosing to share it.
How Locke helps
Locke detects and masks sensitive fields on your device before content is sent to Gemini. Locke — the Sonomos desktop app, coming soon — provides this coverage system-wide, across Google's apps and every other AI surface you use, rather than depending on any single integration.
Keep using Gemini — without the exposure
Locke runs entirely on your device. Sensitive data is detected and masked before any prompt is sent, so nothing confidential ever leaves your machine. Pricing for Locke, the desktop app, is coming soon; Canary is free and open source.
