Security and privacy: common questions

Is it safe to give an AI access to my email?

Inbox Triage keeps your passwords out of the chat, needs your approval before anything sends, and lets you undo changes for 7 days. The mail your AI reads goes to your AI provider, so choose one you trust. See the full answer.

What can the AI see?

The mail synced from the inboxes you connect (about 300 of the newest messages per folder, or 40 to start for Gmail), plus anything it finds when it searches. It can also open a message's full text and fetch an attachment when your task needs it. It never sees Spam, Trash, or your passwords.

Do you store my email?

We store a recent window of messages as plain text, encrypted sign-ins, and attachment names. We do not store attachment files or your full mailbox.

Do you train AI on my email?

No. Inbox Triage never uses your mail to train or improve a language model or any AI app, ours or anyone else's, and does not run an AI model on your mail. The spam filter learns from your own decisions using word-pair counts, not message text.

Can my AI send email without me?

Not unless you turn on Send on my behalf in Settings. That lets an agent you pick send replies to people you have already emailed, after a hold time you set. You can cancel a held send, and you get an alert about it if you turned alerts on. Everything else waits for your approval, and a yes in chat never counts.

Can Inbox Triage staff read my email?

Only if you ask us to look at a specific message, it is needed for security or to investigate abuse, or the law requires it.

How do I stop an AI app from reading my email?

In Settings, under Connected agents, choose Disconnect next to the app. Its access ends right away. For an app that uses an API key, revoke the key under Advanced.

Is Inbox Triage SOC 2 or HIPAA compliant?

No. We do not hold SOC 2 today, and HIPAA is out of scope. We publish the controls we ship on the Security page.