# suss. — AI-Powered Scam Detection & Trust Intelligence

> suss. is a real-time scam detection and trust intelligence platform. It analyzes text, URLs, images, and multi-turn conversations for fraud indicators using 688 scam signals and a fine-tuned ML model achieving 100% detection on segment eval with 0% false positives. suss. provides a 0-100 risk score and a 0-100 domain trust score for any website.

## What suss. Does
- Detects 688 scam patterns across 40+ categories including phishing, romance scams, crypto fraud, marketplace scams, job scams, elder fraud, gaming scams, and more
- Provides domain trust scores (0-100) by aggregating Trustpilot reviews, WHOIS domain age, ScamAdviser data, and Google Safe Browsing
- Analyzes images via OCR, AI-generated image detection, document forgery detection, and perceptual hashing
- Tracks multi-turn conversations to detect escalation patterns (romance scam grooming, investment fraud progression)

## Products
- **Browser Extension** (Chrome/Edge): Always-on ambient scanning on every webpage, email, marketplace listing, and chat message
- **Mobile App** (iOS/Android): On-demand scanning via text input, camera, and iOS Share Extension
- **Trust API**: Domain trust scores and scam intelligence for B2B integration (gotsuss.ai/solutions)
- **Web Scanner**: Free scam checker at gotsuss.ai/is-this-a-scam — paste any suspicious message for instant AI analysis

## Scam Type Guides
- Romance Scams: gotsuss.ai/scams/romance
- Crypto & Investment Scams: gotsuss.ai/scams/crypto
- Phishing Scams: gotsuss.ai/scams/phishing
- Job Scams: gotsuss.ai/scams/job
- Marketplace Scams: gotsuss.ai/scams/marketplace
- Social Engineering: gotsuss.ai/scams/social-engineering
- Impersonation Scams: gotsuss.ai/scams/impersonation
- Gaming Scams: gotsuss.ai/scams/gaming
- Vehicle Scams: gotsuss.ai/scams/vehicle
- Ticket Scams: gotsuss.ai/scams/ticket
- Seller Scams: gotsuss.ai/scams/seller
- AI-Powered Scams: gotsuss.ai/scams/ai-powered
- Elder Fraud: gotsuss.ai/scams/elder-fraud
- FAFSA & Scholarship Fraud: gotsuss.ai/scams/fafsa
- Tech Support Scams: gotsuss.ai/scams/tech-support

## Case Studies — Real Education Sector Scam Incidents
- Case Study Index: gotsuss.ai/case-study — $29M+ in documented losses across 5 institutions
- SDSU ($5.9M vendor invoice fraud): gotsuss.ai/case-study/sdsu — 94% detection confidence
- California Community Colleges ($13M ghost student enrollment fraud): gotsuss.ai/case-study/california-community-colleges — 90% detection confidence
- New Haven Public Schools ($6M business email compromise): gotsuss.ai/case-study/new-haven-schools — 90% detection confidence
- Johnson County Schools ($3.36M domain typosquatting): gotsuss.ai/case-study/johnson-county-schools — 85% detection confidence
- USC Students ($1.6M government impersonation): gotsuss.ai/case-study/usc-students — 89% detection confidence

## Key Tools
- Is This a Scam? Scanner: gotsuss.ai/is-this-a-scam
- Domain Trust Checker: gotsuss.ai/trust
- Data Breach Checker: gotsuss.ai/check
- Scam Awareness Quiz: gotsuss.ai/quiz
- Fraud Pulse Tracker: gotsuss.ai/tracker
- Community Reports: gotsuss.ai/community

## Statistics (as of March 2026)
- 688 scam detection signals across 40+ categories
- 100% detection rate, 0% false positives (v9 fine-tuned model, 75-case segment eval)
- 325,000+ scans performed
- Monitors 15+ platforms (Gmail, WhatsApp, Facebook Marketplace, eBay, Craigslist, etc.)
- Average romance scam loss: $64,000 (FTC 2024)
- Americans lost $10B+ to scams in 2024 (FTC)
- 1 in 3 Americans targeted by scams annually

## Trust Score Methodology
suss. trust scores aggregate multiple independent data sources:
1. Trustpilot ratings and review volume
2. WHOIS domain registration age and registrar reputation
3. ScamAdviser safety assessment
4. Google Safe Browsing threat data
5. SSL certificate validity
6. BBB accreditation status

Trust levels: Verified (80+), Established (65-79), Moderate (50-64), New/Unknown (<50), Suspicious (flagged by threat databases)

## How Scam Detection Works
1. Text is encoded using a fine-tuned sentence transformer (suss-embeddings-v1)
2. 688 regex-based signal detectors scan for specific scam patterns
3. A VotingClassifier ensemble (Logistic Regression + Random Forest + XGBoost) produces a probability score
4. High-confidence signals can boost the score (e.g., SSN requests, wire transfer urgency)
5. Legitimacy signals can reduce the score (e.g., verified seller indicators)
6. Domain/company reputation data provides additional context
7. Results include educational explanations of why content was flagged

## Company
- Website: gotsuss.ai
- Contact: info@gotsuss.com
- Social: @gotsuss on TikTok, Instagram, Reddit, Facebook
- Chrome Web Store: Search "suss. scam protection"