Privacy & Data
Everything about how MailPrism learns from you, handles images and trackers, and stores your data lives under Settings → Privacy. Each control has a sensible default — you only need to change the ones that matter to you.
This page does not cover the AI Data Processing consent switch — the master on/off for all AI features. See AI Privacy & Consent for that.
Pattern learning
MailPrism can learn from how you organize your inbox in Gmail, so it can suggest rules and improve over time. The Pattern Learning card controls exactly what it watches.
Enable Pattern Learning is the master toggle (on by default — it's a core feature). When it's on, you choose which actions to learn from:
| Setting | Learns from | Default |
|---|---|---|
| Learn from Archives | When you archive emails | On |
| Learn from Labels | When you apply or remove labels | On |
| Learn from Stars | When you star or unstar emails | On |
| Learn from Deletes | When you delete emails (limited usefulness) | Off |
Turn off any signal you'd rather MailPrism ignore. Turning the master toggle off stops all of this learning at once.
→ More on what gets learned and how corrections work: Pattern learning
Classification feedback
The Classification Feedback card links to Settings → Privacy → AI corrections
(/settings/privacy/ai-corrections). There you can review the corrections you've given
the AI about how it classified your email — and undo or delete them.
→ See your suggestions: Rule suggestions
Reading behavior
The Reading Behavior card controls what happens to an email when you open it.
- Auto-Mark as Read — marks an email as read a short time after you start viewing it. Off by default — opening an email leaves it unread until you turn this on.
- Mark as Read Delay — when auto-mark is on, choose how long to wait: 1, 2, 3 (default), 5, or 10 seconds.
A longer delay means quick glances won't mark a message read by accident — handy if you scan your inbox before deciding what to act on.
Email privacy
The Email Privacy card controls how images and hidden trackers are handled when you read mail.
| Setting | What it does | Default |
|---|---|---|
| Block All External Images | Replaces every image with a placeholder until you choose to load it | Off |
| Block Tracking Pixels | Detects and blocks hidden trackers while still showing normal images | On (Recommended) |
| Show Blocked Image Warnings | Shows a banner when images are blocked, with a button to load them | On |
A few notes:
- Block Tracking Pixels is the recommended middle ground — you keep normal images but hidden trackers are stripped.
- When Block All External Images is on, the tracking-pixel toggle is disabled, because everything is already blocked.
- Proxy Images (loading images through MailPrism to hide your IP) is shown as Coming Soon and not yet available.
Activity log
The Activity Log card shows recent email actions MailPrism has tracked for pattern learning. As you organize emails in Gmail, those actions appear here.
- Items show the action, details, and how long ago it happened.
- Auto-tracked items (from watching your Gmail) are marked as such.
- If pattern learning is off, the log invites you to turn it on to start tracking.
How long it's kept: the activity log follows your plan's history length, with a minimum of 90 days — today that means 90 days. The Data Retention setting below doesn't change it.
Advanced settings
The Advanced Settings card sets how long history and logs are kept.
- Data Retention — how long MailPrism keeps your rule execution history, AI usage logs, and ClickUp forwarding logs before automatic cleanup: 30, 60, 90 days (default), 180 days, or 1 year.
The activity log isn't affected — it follows your plan (see above).
Your data
The Data Management card handles exporting your data and clearing what MailPrism has learned, in line with your data-protection rights (GDPR).
Export your data
Export Data downloads a copy of all your account data as a JSON file — settings, rules, activity, and learning data. The date of your last export is shown for reference.
Clear learning data
Clear AI Learning Data permanently deletes what MailPrism has learned from your email and activity. The confirmation dialog lists exactly what happens:
| Examples | |
|---|---|
| Deleted | Detected patterns, pending suggestions, learned writing data (analysed sent emails, unreviewed greetings and closings, edit patterns), saved training replies, contact relationships MailPrism worked out on its own |
| Kept, but emptied | Writing profiles keep their names, instructions, and settings, but lose learned tone, formality, length, and samples |
| Not touched | Your activity log (reply tracking relies on it), your sender corrections, accepted or dismissed suggestions and their rules, your settings, contacts, and account |
MailPrism starts learning again from your new activity, so suggestions and drafts may be less personal for a while.
Cleared learning data can't be recovered. Export first if you want a copy.
Account deletion
Deleting your whole account (not just learning data) is handled elsewhere. See Account & Security → Delete your account.
Related
- AI Privacy & Consent — the master AI switch.
- Pattern learning — what MailPrism learns and how.
- Privacy & Security — the broader picture.
- Account & Security — profile, sign-in, and account deletion.