What is ScamFlag?
ScamFlag is an AI-powered real-time, omnichannel scam detection that protects users and banks’ reputations. It analyzes images—including photos and screenshots—from any digital channel (SMS, WhatsApp, email, web) to detect deceptive content, phishing attempts, fake payment requests, and fraudulent sites.
With cross-channel coverage and bank-grade signal sharing, ScamFlag helps your institution recognize threats early and intervene before scams turn into fraud.
Simple 3-Step Use
1. Open the app
2. Take/upload a screenshot
3. Get a scam risk analysis and recommended actions.
Secure
Data is securely stored and auto-deleted after a set time.
Fast and accurate
Get instant results with 99% detection accuracy.
Omnichannel Protection
Identifies scams across all digital channels—spotting signs of investment fraud, romance scams, smishing, phishing, and more.
Any questions?
Check out the FAQs
The ScamFlag is designed to detect symptoms of common tricks fraudsters, particularly scammers, use to trick users. These include (but are not limited to) marketplace scams, romance scams, investment scams, fake merchant scams, phishing, smishing and business email compromise.
It works across multiple channels and modalities, so users can take or upload pictures of websites advertising products, bank login screens, WhatsApp and SMS conversations, e-mail messages, and more.
Users benefit from real-time responses indicating whether what they see is likely a scam or other form of fraud attempt, including recommended steps to take.
ScamFlag is based on a GenAI agent trained on a large number of scam samples to analyze omnichannel information for signs of suspicious text strings, deceptive language, and fake graphical artifacts. It analyzes any picture from the structural perspective, extracts visible text found in the picture, and tries to understand the context. If it finds any link or bank account number, it checks it against public sources and the ThreatMark internal intelligence database.
The detection efficacy of ScamFlag depends on the quality of the data provided. When evidence is submitted — for example, a screenshot of the conversation or a photo of the phishing website — ScamFlag achieves an accuracy of approximately 99% compared to a live fraud expert. In other words, 99 out of 100 submitted screenshots or photos are correctly evaluated.
Based on the provided evidence — such as a screenshot of a conversation or a photo of a phishing website — ScamFlag will deliver a clear result along with a detailed explanation. For example, if a phishing e-mail is submitted, ScamFlag will list all the attributes that led to its classification as phishing (e.g., unsolicited links, urgency in the message, unknown sender, untrusted domain, and other indicators).
Typically not. It will use permissions already granted to the hosting application — for example, banking apps generally have access to storage and the camera (only those needed) by default.
No—ScamFlag is seamlessly integrated into your existing banking app, ensuring a frictionless user experience.
ScamFlag is priced based on the size of your user base, with annual licensing starting in the lower tens of thousands of dollars.
Yes—ScamFlag works as a standalone solution and can easily complement your existing tech stack. However, for the best results, we recommend using it as part of the full ThreatMark platform.
Test it right away!
Sing up to get access to the ScamFlag API and sample code
Experience the power of real-time scam detection by test-driving the ScamFlag API today. Whether you're evaluating new fraud prevention capabilities or looking to enhance your own app, the ScamFlag API lets you instantly analyze images and text for signs of scams.
All seamlessly integrated into your banking app — ScamFlag is fully customizable and can be white-labeled to match your bank’s brand.
Start your integration now and see just how simple, accurate, and effective scam detection can be.