ScamShield AI

AI-powered scam detection platform for messages, suspicious links, and screenshots.

Comment

Inspiration

Online scams, phishing links, and fraudulent messages are becoming increasingly common. Many users struggle to identify scams before becoming victims. We wanted to create a simple AI-powered tool that helps users quickly determine whether a message, URL, or screenshot is suspicious.

What it does

ScamShield AI analyzes suspicious content using Google Gemini AI and provides:

  • Scam detection for messages
  • URL risk analysis
  • Screenshot scam detection
  • Risk score assessment
  • AI-generated explanation and recommendations

How we built it

We built ScamShield AI using:

  • React + Vite for the frontend
  • Flask for the backend
  • Google Gemini AI for content analysis
  • Flask-CORS for frontend/backend communication
  • Pillow for image processing

The application sends user input to the Flask backend, where Gemini AI evaluates the content and returns a risk assessment with an explanation.

Challenges we ran into

  • Integrating AI analysis reliably
  • Handling API key security
  • Connecting React frontend with Flask backend
  • Managing screenshot processing and analysis

Accomplishments that we're proud of

  • Built a working AI-powered scam detection platform
  • Implemented message, URL, and screenshot analysis
  • Created a clean and user-friendly interface
  • Secured API key handling using environment variables

What we learned

  • Full-stack AI application development
  • Secure API key management
  • AI prompt engineering
  • Frontend-backend integration

What's next for ScamShield AI

  • Browser extension
  • WhatsApp scam detection
  • Email phishing analysis
  • Mobile application
  • Real-time website scanning
  • Multi-language support
Share this project:

Updates

Submission history