Skip to content

The Memory privacy commitment

Memory never phones home.

This is the load-bearing claim that separates Memory from every other "AI for your browser." It is not marketing copy — it is a constraint we hold ourselves to in the code.

What stays on your device

  • Your captured pages
  • Your queries against those captures
  • The embeddings Memory uses to search them
  • The AI model itself (Gemma 4, running locally in your browser via WebAssembly)
  • Every Memory chat session, when chat unlocks

What crosses the network

A short, fixed list:

  1. Authentication. When you sign in, the Sauvrn web app talks to our auth provider (WorkOS). No browsing data is involved.
  2. License verification. Every six hours, the extension calls our backend to confirm your license is still active. The request contains your license token. It does not contain anything about what you've read or asked.
  3. Model artifacts. The first time you launch Memory, the extension downloads the Gemma 4 model weights and the WASM runtime from models.sauvrn.ai. After that, the model runs locally with no further network calls.
  4. Opt-in feedback. If you click "Send feedback" inside the extension and type a message, that message is sent to us. We sanitize it before it lands in our system to scrub any tokens you may have pasted accidentally.

That's the entire list. If you ever see network activity from Memory that doesn't match the list above, that's a bug — please report it via the in-extension feedback form, or email support@sauvrn.ai.

Why we made this constraint

Sending browsing data to a server "for AI" is a one-way trip. Once your reading lands somewhere we host, it sits in logs, backups, and shared dashboards forever. Even with the best intentions, that data can leak. Memory is built so the leak surface does not exist.

It also means we can never read your data ourselves, even if we wanted to. We made that trade on purpose.

Open questions we're still working on

  • Cross-device sync. Today, Memory's index is per-device. Syncing without breaking the privacy commitment is technically hard. We're working on it.
  • Workspace sharing. Team plans will eventually let you opt in to sharing specific pages or answers with your team. That sharing will be page-by-page and explicit — not bulk telemetry.

Audit it yourself

Memory's source is closed, but the network behavior is auditable from any browser's developer tools. Watch the Network tab while you use chat. You'll see calls to api.sauvrn.ai (license verification, every six hours), models.sauvrn.ai (only on first launch), and nothing else. If we ever break that, you'll see it before we do.