What is Email Triage AI?

Email Triage AI is a web app that connects to your Gmail or Outlook inbox and uses AI to read incoming emails, categorise them, score their priority, and draft replies for the ones that need a response. It sits alongside your existing email client rather than replacing it.

Which email providers does Email Triage AI support?

Gmail and Outlook are both supported. Connection takes under a minute via standard OAuth — no app passwords, no IMAP credentials.

How does Email Triage AI handle my email privacy?

Emails are processed to generate categorisation, priority scores, and drafts. We do not sell or share email content with third parties. See the privacy and transparency pages for the full data-handling description.

Is there a free trial?

Yes — the pricing page lists the current free tier and paid plans. You can connect Gmail or Outlook and try the categorisation and draft features without entering payment details.

Does Email Triage AI send replies automatically?

No. Email Triage AI drafts replies for emails that need one and surfaces them for review. You stay in control of every send.

Privacy Policy — Email Triage AI

Last updated: 18 April 2026

1. Introduction

Email Triage AI ("we", "our", or "us") is committed to protecting your privacy. This Privacy Policy explains how we collect, use, disclose, and safeguard your information when you use our email triage application.

2. Information We Collect

Account Information

Email address
Name (from your OAuth provider)
Authentication tokens (encrypted at rest with AES-256-GCM)

Email Content and Metadata

To categorise and draft replies to your emails, we read and store:

Sender name and email address
Subject lines
Received timestamps
The first 10,000 characters of the email body (truncated). This is stored in our database so we can display the email in your dashboard, generate draft responses, and let you correct AI categorisations later.
AI-generated category, priority score, and draft response

We do not store attachments. Email content is only accessible to you and to our processing pipeline; it is never sold, never shared with third parties for advertising, and never used to train third-party AI models.

Bring-your-own-key (optional)

You can supply your own Anthropic API key in Settings. When set, all AI calls for your account are sent using your key, so Anthropic bills the calls to your account, not ours. The key is encrypted at rest with AES-256-GCM (the same cipher we use for OAuth tokens), is only decrypted in memory at the moment of use, and is never logged or transmitted to the frontend (Settings shows only a masked preview like sk-a…XYZ4). You can remove the key at any time, after which billing reverts to us.

Follow-up reminders

When you open the Follow-ups page, we query your Gmail "Sent" folder to find messages older than the threshold you've set (default 3 days) where the most recent message in the thread is still from you (i.e. nobody has replied). We read message metadata (subject, recipient, send date) and check thread state. The only thing persisted is the thread IDs you have dismissed ornudged, so we don't re-surface them.

Writing-style learning (optional)

If you opt in via Settings → "Learn my style", we read your last ~20 sent emails to generate a short summary of your writing voice (typical greeting, length, tone, sign-off). Only the AI-generated summary is stored — the raw sent-mail content is never persisted. The summary is then included when AI drafts replies on your behalf so they sound like you. You can clear it at any time from Settings.

3. How We Use Your Information

To provide email categorisation and prioritisation
To generate AI-powered response drafts
To improve the accuracy of our categorisation models from your corrections
To send notifications about high-priority emails (with your consent)

4. Data Security

We implement industry-standard security measures including:

Encryption of OAuth tokens at rest using AES-256-GCM
HTTPS (TLS 1.2+) for all data in transit
Secure authentication via Google and Microsoft OAuth (we never see your password)
Strict-Transport-Security, Content-Security-Policy, and X-Frame-Options headers
Regular security audits and dependency updates

5. Third-Party Services

We use the following third-party services:

Google OAuth and Gmail API — to read and authenticate your Gmail account
Microsoft OAuth and Graph API — to read and authenticate your Outlook account
Anthropic Claude — for AI categorisation and draft generation. Email content sent for processing is governed by Anthropic's zero-retention enterprise terms.
Stripe — for subscription billing (we never store your card details)
Cloudflare — for content delivery, DDoS protection, and basic analytics

6. Your Rights

Under GDPR, CCPA, and similar regimes you have the right to:

Access a copy of your personal data
Request deletion of your data and account
Disconnect your email accounts at any time (this also revokes our Google/Microsoft tokens)
Export your stored email categorisations and drafts
Lodge a complaint with your local data protection authority

7. Data Retention

Stored email metadata, body text, and AI outputs are retained for as long as your account is active. When you delete your account or disconnect a mailbox, associated data is permanently removed within 30 days.

8. Cookies and Tracking

We use a small number of strictly necessary cookies to keep you signed in. We use Cloudflare Web Analytics, which is cookie-less and does not fingerprint visitors. We do not use third-party advertising trackers.

9. Contact Us

For questions, data access, or deletion requests, use our contact form (Privacy topic).