01
Open a supported message
ScamBrake looks only at the active incoming message—not an entire mailbox or page.
Local checks, cloud by choice
See the warning signs before you act.
ScamBrake checks supported incoming messages in your browser and explains why something looks suspicious. Message content stays on your device unless you explicitly request an AI second opinion.
No account requiredNo message storageOptions start off
Example message
Delivery Support
Action required: delivery fee
Your parcel is being held. Pay the small release fee today to prevent it from being returned.
Potential delivery-fee scam
High riskThis message combines urgency, a payment request, and a link you were not expecting.
Urgent deadlineUnexpected feeSuspicious link
Checked locallyNo message stored
Illustrative interface. No real message data.
Local by default
Core checks happen in your browser. Normal scans do not upload message content.
AI only when requested
You see the masked text before an optional AI second opinion is sent.
Sharing is separate
Cloud history and caregiver sharing stay off until you enable each one.
Current coverage
Clear about where it works.
Provider interfaces change frequently. These labels describe the current extension build; live browser acceptance is completed for each release before a provider is promoted beyond beta.
GmailCore beta
OutlookCore beta
Yahoo / AOLTesting
Proton MailTesting
WhatsApp WebTesting
MessengerTesting
Facebook postsManual check
How it works
A warning, not a black box.
02
Review the warning signs
A local check highlights suspicious requests, pressure tactics, and risky payment language.
03
Choose what happens next
Stop, dismiss the warning, or explicitly request an optional masked AI second opinion.
Privacy in practice
Your message is not the product.
ScamBrake separates local detection, optional AI, cloud history, caregiver sharing, and analytics into distinct choices. Turning on one does not silently enable the others.
Local scan
The active message is analyzed in your browser.
Optional AI
Masked text is shown to you before it is sent.
Optional history
Only derived warning metadata can be synced.
Active message
Stays in the browser
Local warning
Risk, category, and understandable reasons
Optional AI second opinion
Only after preview and confirmation
Made for real decisions
Warnings without taking away control.
For anyone checking a suspicious message
Plain-language warnings help you pause, recognize pressure tactics, and decide what to do next.
- Prominent warnings
- Reasons you can understand
- No account required for local checks
For families who choose to help
Caregiver sharing is permission-based and limited to derived warning metadata—never message bodies.
- Separate opt-in
- High-risk metadata only
- Control stays with the protected person
Start with private, local warnings.
Install the extension, keep every optional data flow off, and enable only the features you choose.
Get ScamBrake