How it works
How Does AI Email Scanning Work? From Inbox to Verdict in Under a Minute
You don't need to install anything, switch email clients, or learn a new tool. If an email looks suspicious, just forward it like you would any other email. ScamDrill takes it from there.
You forward the suspicious email
From Gmail, Outlook, Apple Mail, or any other email app, forward the message to check@scamdrill.com from the email address attached to your ScamDrill account. We recognize you from your sender address — no login or extra setup needed.
ScamDrill parses the full message
Our system extracts the sender headers, subject, body text, and any links or attachments. We check what the email really looks like under the hood — not just what's displayed — so spoofed sender names and hidden link targets are caught.
AI analyzes content, context, and intent
A large language model trained on real phishing patterns evaluates the message the way an experienced security analyst would: Is the sender really who they claim to be? Is the request plausible? Do the urgency cues, wording, and links match a known scam playbook?
You get a verdict and an explanation
Within seconds you receive a reply with a clear verdict, a risk score, the specific red flags spotted, and plain-English guidance on what to do next — whether that's delete it, report it, or treat it as legitimate.
What it checks
Six layers of inspection, every time
Good pattern matching can catch sloppy scams. Modern AI-generated phishing is often pixel-perfect and grammatically clean — so we inspect the signals that still give them away.
Sender authenticity
Checks SPF, DKIM, and DMARC records, lookalike domains, and display-name spoofing.
Link safety
Unfolds shortened URLs, compares displayed vs. actual destinations, and flags newly registered or impersonating domains.
Content & tone
Reads the body for urgency cues, manipulation tactics, and language consistent with known scam categories.
Brand impersonation
Compares claimed senders (banks, IRS, Amazon, DHL…) against how those brands actually communicate.
Known-scam playbook match
Matches against patterns from the FTC, FBI, and our own simulation library — IRS refunds, package redelivery, tech support, grandchild, etc.
Attachment review
Flags common malware carriers (.zip, .html, .iso, macro-enabled Office docs) without ever executing them.
Example
What a reply actually looks like
Here's a real-world example of the kind of message people forward to check@scamdrill.com and the structured analysis that comes back.
The forwarded email
A text message or email along these lines — professionally written, just urgent enough, with a short link.
Verdict: Do not click the link. This looks like a package-redelivery phishing scam designed to capture payment details.
Why we flagged it
- Sender domain "usps-alerts.net" is not an official USPS domain (usps.com).
- USPS does not charge "redelivery fees" by email or text.
- Link destination "usps-redeliver.co" was registered recently and is unrelated to USPS.
- Urgency cue ("within 12 hours") is a common scam pressure tactic.
What to do
Delete the email. If you're expecting a package, check its status directly at usps.com using the tracking number from your original order.
Why pair scanning with simulated training?
Simulated scam drills train the instinct — the gut feeling that something's off. AI email scanning gives that instinct somewhere to turn when it fires. Together they form a habit loop: practice, notice, verify.
- Training builds pattern recognition over weeks and months of spaced-out simulations.
- Scanning provides instant, authoritative answers when a real email feels suspicious.
- Everything a learner forwards also becomes data the guardian sees in the dashboard, so you know what they're flagging.
Privacy & limits
What we do — and don't — do with your email
We only read what you forward
ScamDrill never touches your inbox. The only messages we see are the ones you deliberately forward to check@scamdrill.com.
We don't train on your content
Forwarded messages are used to generate your analysis and are retained only for as long as needed to show it in your dashboard.
Links are never "opened"
URLs are inspected via metadata and reputation signals — our system doesn't fetch or render pages on your behalf.
Fair-use limits
Up to 10 checks per day per learner. That's enough for real suspicious messages without opening the door to abuse.
Pricing
Included with the Family+ plan and with Business and Enterprise organization plans — no add-on purchase required.
On Free and Family plans, AI Email Scanning is available as an add-on at $2.99/mo or $29.99/yr (save 17%).
You can bundle the add-on when you sign up, or add it later from your billing dashboard. Cancel anytime from the Stripe billing portal.
AI Email Scanning: Frequently Asked Questions
No. AI email scanning works through plain email forwarding, so it's compatible with every major email provider and device. Forward to check@scamdrill.com and the analysis comes back to the address you forwarded from.
Most analyses are returned in under 30 seconds. Complex messages with many links or large attachments can take up to a minute.
Yes. Every learner attached to your ScamDrill account can forward messages to check@scamdrill.com from the email address associated with their profile. Each learner has their own 10-check daily allowance.
You can paste the content of a suspicious text message into an email and forward it just like any other message. SMS-native submission is on our roadmap for Pro plans.
No — think of it as a second opinion. Spam filters decide automatically. AI email scanning is for the messages that get through your filter and still look questionable, where you want a real explanation instead of a binary decision.
Every analysis includes the reasoning behind the verdict so you can evaluate it yourself. You can flag any result in your dashboard and we use that feedback to continually improve detection quality.