Safety first

Is this safe?

Worried about privacy or breaking something? Good. That means you're careful. Here are my plain answers, gathered from every guide.

Milo's 5 golden rules

The postcard rule. If you wouldn't write it on a postcard, don't paste it in.
Keep "ask me first" on. Let AI suggest. You approve.
Start in a test folder. Never hand over your whole computer on day one.
Check its work. AI can sound sure and still be wrong. Look before you send.
Connect one app at a time. Add more only when you trust it.

Connectors let Muse see your calendar, write emails, and more. Start with just one.

Safety facts, straight from Meta:

By default, Muse won't do many big things, like sending an email, without your okay. You can change when it asks in Settings.²
Many connectors can be set to look only. Muse can read but not act. That's the best way to start.²
Connecting email lets Muse search it. It won't download your whole inbox unless you ask.²
Disconnect any time in Settings → Connectors. Muse may still remember what it already learned, but you can ask it to forget.²
Custom connectors (for apps not on the list) aren't checked by Meta. Beginners should skip them.²

Three cool ideas you can have your AI do, in your inbox every morning. Tested first, no hype, no homework. Unsubscribe any time.

We never sell your email. See our privacy policy.