🔍 Inspiration With the rapid rise in AI-generated content and deepfake technologies, online identities are becoming increasingly vulnerable. Scams, impersonations, and fake job offers on platforms like LinkedIn and Instagram are surging. We envisioned Cryptex ID as a real-time AI-powered identity verification system to safeguard users from falling victim to deceptive profiles. Our aim is to combine AI and Blockchain to restore trust and transparency in digital interactions.
🧠 What It Does Cryptex ID detects fake, AI-generated, or suspicious profiles across social platforms and job portals. Here's how it works:
🔍 Image Analysis using a DeepFake CNN model to detect altered or AI-generated profile photos.
✍️ Text Analysis using GPT-4 to evaluate the authenticity of bio, posts, or job offers.
🔗 Blockchain Logging to maintain a tamper-proof identity validation ledger.
🧩 Chrome Extension UI to provide real-time alerts and verification badges.
🏗️ How We Built It Frontend: React.js + Chrome Extension
Backend: Node.js with Express
Database: MongoDB Atlas
AI Models:
DeepFake detection CNN (image-based)
GPT-4 (text pattern & context validation)
Blockchain: Identity data hashed and logged on Ethereum testnet
APIs: Instagram, LinkedIn, Twitter scraping with user consent
🚧 Challenges We Faced Building a lightweight deepfake detection model that can run fast enough in real-time.
Balancing accuracy vs false positives while verifying text and images.
Integrating blockchain logging without slowing down performance.
Chrome Extension integration and UX testing across multiple platforms.
✅ Accomplishments Developed a functional identity checker that flags fake profiles with >85% accuracy.
Seamless integration of AI + Blockchain with a user-friendly Chrome interface.
Created a secure backend to handle AI results and log data.
📚 What We Learned Practical deployment of AI models in real-time applications.
Secure identity handling using blockchain.
Browser extension development with seamless backend-AI integration.
🌱 What's Next Support for more platforms (Facebook, WhatsApp Business).
Improve deepfake model with transfer learning from larger datasets.
Incorporate voice and video authentication for deeper checks.
Expand into KYC and recruitment platforms for enterprise adoption.
Built With
- chrome-extension-apis
- deepfake-cnn-model
- express.js
- instagram-graph-api
- javascript
- mongodb
- node.js
- openai-gpt-4-api
- react
- tailwind-css
- tensorflow
Log in or sign up for Devpost to join the conversation.