Threat detection
Real threats found. Noise filtered out.
Unphish runs every signal through five layers of detection, combining custom rules, machine learning, AI reasoning and a final check by an analyst. You get confirmed phishing sites, scams and impersonation, including the ones built to hide from scanners, without the false alarms.
How it works
Five layers between a signal and your feed.
Each layer narrows thousands of raw signals down to the few that matter. Automation does the heavy lifting, and a person makes the final call.
The results
Accurate enough to act on.
From cases we detected and closed for our clients in the first half of 2026.
What's included
Everything you need to find threats early.
FAQ
Threat detection questions, answered.
What is threat detection?
It's the process of deciding which of the things found by monitoring are real threats to your brand. Unphish takes in signals from domains, websites, ads, social media, messaging channels and reports, then works out which are phishing sites, scams or impersonation and how urgent they are.
How does the five-layer detection work?
Signals are collected, then filtered with rules built for your brand. Machine learning checks page code, logos and visuals. AI-assisted analysis reads context and intent and gives a risk score. Finally, an analyst confirms the threat before it reaches your feed or goes to takedown.
How do you avoid false positives?
Every threat is checked by an analyst before any takedown, and your official sites and accounts are whitelisted. It shows in the results. In the first half of 2026, fewer than 1% of our reports were rejected by the platform they were sent to.
Is it all done by AI?
No. Automation and AI-assisted analysis do the heavy lifting, but a person makes the final call. The AI risk score sets priority and never filters a threat out, and analyst decisions feed back into the models so detection keeps improving.
Can you detect scam sites that hide from scanners?
Yes. Some scam sites only show the scam on a phone, or redirect reviewers somewhere harmless. We load pages the way customers see them, so the hidden content is caught and captured as evidence. For one airline group, every one of the more than 180 mobile-only scam sites we pursued was taken down.
What happens after a threat is detected?
It appears in your threat feed with evidence, a risk score and its full detection history. From there it can go to enforcement in one click, where configured. In the first half of 2026, 96% of the threats we pursued were taken down.
Does Unphish help with the FTC Impersonation Rule?
Yes. The FTC Impersonation Rule, in force since April 2024, allows civil penalties up to $53,088 per violation for impersonating a business. Unphish finds and takes down those impersonations fast, with a complete audit trail of every detection and takedown.
See where your name is being misused.
Run a free threat scan or book a demo to see what's targeting your customers right now.