Our mission
Scams have become industrial. Pig-butchering rings, fake delivery SMS, romance fraud, phishing kits and impersonation attacks now cost ordinary people more money each year than burglary, car theft and pickpocketing combined. ScamGuard exists to give every person on the internet a free, neutral, AI-powered second opinion before they click, pay or trust.
What we do
- AI scam checking. Paste a link, message, screenshot, voice note or phone number and our models return a verdict in seconds.
- Community scam database. A public, searchable archive of reported scam messages, phone numbers, websites and profiles.
- Subscription Shield. Bank-statement scanning that surfaces forgotten subscriptions and silent free-trial conversions.
- Gmail Shield. Inbox scanning that flags phishing, breach notifications and predatory recurring charges.
- WhatsApp bot. The same scam detection, on WhatsApp, with no app install required.
How ScamGuard works
Every check combines four signal layers: large-language-model analysis of the text or image, a curated database of known scam infrastructure (phone numbers, wallets, domains, profiles), community reports submitted by ScamGuard users worldwide, and open-source threat intelligence. Verdicts are explained in plain language, with the specific red flags the model found.
Who we are
ScamGuard is built by the ScamGuard Security Research Team — a small, independent group of engineers and analysts focused on consumer-facing fraud. We publish our editorial standards and our fact-checking process publicly so anyone can see how a verdict is reached.
How we make money
Most ScamGuard features are free, with usage-based credits for heavier use and Subscription Shield premium tools. We do not sell user data, we do not run third-party advertising on scan results, and our verdicts are never influenced by paid placement.
Contact and corrections
If you believe ScamGuard has misclassified a website, profile or message, please use our contact page to request a review. We publish corrections per our editorial policy.