AI-Powered Fake News Detection & Claim Verification System
π This repository contains NO real credentials. All sensitive data uses placeholders.
π New users: Read SETUP_SECRETS_GUIDE.md to configure your own credentials.
Before contributing or committing:
- β Read SECURITY_README.md
- β Check DELETE_BEFORE_COMMIT.md
- β
Never commit
.envfiles with real credentials
ZeroTRUST is an advanced AI-powered platform that helps combat misinformation by:
- π Automated Fact-Checking: Analyzes claims against multiple trustworthy sources
- π€ Multi-Agent AI System: Uses specialized AI agents (Research, News, Scientific, Social Media)
- π₯ Media Analysis: Detects deepfakes, manipulated images, and audio
- π Real-time Verification: Provides credibility scores and evidence trails
- π Multi-Platform: Web portal, browser extension, and API access
- Architecture
- Features
- Quick Start
- Setup & Configuration
- Documentation
- Technology Stack
- Contributing
- License
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β User Interfaces β
β Web Portal β Browser Extension β Mobile App (Future) β
βββββββββββββββββββ¬ββββββββββββββββββββββββββββββββββββββββββββ
β
βΌ
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β API Gateway (Node.js) β
β Authentication β Rate Limiting β Request Routing β
βββββββββββββββ¬ββββββββββββββββββββββββββββββββ¬ββββββββββββββββ
β β
βΌ βΌ
βββββββββββββββββββββββββββββββ ββββββββββββββββββββββββββββ
β Verification Engine (Python)β β Media Analysis (Python) β
β βββββββββββββββββββββββββ β β ββββββββββββββββββββ β
β β Manager Agent (LLM) β β β β Image Analysis β β
β β ββ Research Agent β β β β Video Analysis β β
β β ββ News Agent β β β β Audio Analysis β β
β β ββ Scientific Agent β β β β Text Extraction β β
β β ββ Social Media Agent β β β ββββββββββββββββββββ β
β βββββββββββββββββββββββββ β ββββββββββββββββββββββββββββ
βββββββββββββββ¬ββββββββββββββββ
β
βΌ
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
β AWS Services β
β Bedrock (LLM) β S3 β Lambda β DynamoDB β CloudWatch β
βββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββββ
- β Multi-Source Verification: Checks claims against news APIs, academic papers, fact-check databases
- β AI Agent Orchestration: Manager agent coordinates specialized agents for comprehensive analysis
- β Media Forensics: Detects manipulated images, deepfake videos, and synthetic audio
- β Evidence Tracking: Provides source citations and confidence scores
- β Real-time Analysis: Fast processing with caching for repeated queries
- β User Authentication: Secure JWT-based auth system
- β Browser Extension: Verify claims directly while browsing
- π Scalable Architecture: Microservices-based design
- π Comprehensive Logging: CloudWatch integration for monitoring
- π Security First: No hardcoded credentials, JWT authentication
- π³ Docker Support: Easy deployment with Docker Compose
- π§ͺ Extensive Testing: Unit, integration, and E2E tests with Playwright
- π AWS Integration: Leverages Bedrock, S3, Lambda, DynamoDB
- Node.js 18+ and npm
- Python 3.11+
- Docker & Docker Compose (optional)
- AWS Account with Bedrock access
- PostgreSQL database (or Supabase free tier)
git clone https://github.com/your-org/ZeroTRUST-AI4Bharat.git
cd ZeroTRUST-AI4Bharat# Read the setup guide
cat SETUP_SECRETS_GUIDE.md
# Copy environment templates
cp apps/api-gateway/.env.example apps/api-gateway/.env
cp apps/web-portal/.env.example apps/web-portal/.env
# Edit with your credentials
nano apps/api-gateway/.envRequired credentials:
- AWS Access Key ID & Secret
- Database connection string (PostgreSQL/Supabase)
- JWT secret (generate with
openssl rand -hex 64)
See detailed instructions: SETUP_SECRETS_GUIDE.md
# API Gateway
cd apps/api-gateway
npm install
# Web Portal
cd ../web-portal
npm install
# Verification Engine
cd ../verification-engine
python -m venv .venv
.venv\Scripts\Activate.ps1 # Windows
source .venv/bin/activate # Linux/Mac
pip install -r requirements.txtOption A: Using Docker Compose (Recommended)
docker-compose upOption B: Manual Start
# Terminal 1: API Gateway
cd apps/api-gateway
npm run dev
# Terminal 2: Verification Engine
cd apps/verification-engine
uvicorn src.main:app --reload --port 8000
# Terminal 3: Web Portal
cd apps/web-portal
npm run dev
# Terminal 4: Media Analysis
cd apps/media-analysis
uvicorn src.main:app --reload --port 8001- Web Portal: http://localhost:5173
- API Gateway: http://localhost:8000
- API Docs: http://localhost:8000/docs
| Guide | Description |
|---|---|
| SETUP_SECRETS_GUIDE.md | START HERE - Complete credential setup guide |
| SECURITY_README.md | Security best practices and verification |
| DELETE_BEFORE_COMMIT.md | Pre-commit checklist |
| Step | Guide | Description |
|---|---|---|
| 1 | 01-iam-setup.md | IAM user and permissions |
| 2 | 02-bedrock-setup.md | Enable Bedrock models |
| 3 | 03-s3-setup.md | S3 bucket for media |
| 4 | 04-dynamodb-setup.md | DynamoDB tables |
| 5 | 05-cloudwatch-setup.md | CloudWatch logging |
| 6 | 06-lambda-setup.md | Lambda functions |
| 7 | 07-media-services-setup.md | Media analysis services |
- Free Tier Guide: QUICKSTART_FREE_TIER.md
- AWS Migration: AWS_MIGRATION_GUIDE.md
- Testing Guide: TESTING_GUIDE.md
- Architecture Analysis: COMPREHENSIVE_ARCHITECTURE_ANALYSIS.md
- Implementation Status: IMPLEMENTATION_STATUS.md
- Development Plan: DEVELOPMENT_PLAN.md
- Testing Guide: E2E_TESTING_GUIDE.md
- AWS Services Analysis: AWS_SERVICES_USAGE_ANALYSIS.md
- Free Tier Optimization: FREE_TIER_OPTIMIZATION.md
- Deployment Steps: STEP_4_ECS_DEPLOYMENT.md
- React + TypeScript + Vite
- TailwindCSS for styling
- Axios for API calls
- Browser Extension (Vanilla JS)
- Node.js + Express (API Gateway)
- Python + FastAPI (Verification Engine & Media Analysis)
- PostgreSQL (Database)
- Redis (Caching)
- AWS Bedrock (LLM inference)
- Amazon Nova models
- Mistral Pixtral
- LangChain (Agent orchestration)
- Custom AI Agents (Research, News, Scientific, Social)
- Bedrock: LLM hosting
- S3: Media storage
- Lambda: Serverless functions
- DynamoDB: NoSQL cache
- CloudWatch: Logging & monitoring
- Rekognition: Image analysis
- Textract: Text extraction
- Transcribe: Audio transcription
- Docker & Docker Compose
- Playwright (E2E testing)
- Pytest (Python testing)
- Jest (JavaScript testing)
We welcome contributions! Before contributing:
- Read security guidelines: SECURITY_README.md
- Never commit real credentials
- Follow the development plan: DEVELOPMENT_PLAN.md
- Run tests before submitting PRs
# 1. Create a feature branch
git checkout -b feature/your-feature-name
# 2. Make your changes
# 3. Run tests
npm test # Frontend tests
pytest # Python tests
npm run test:e2e # E2E tests
# 4. Verify no secrets
git diff --staged | grep -iE "AKIA|secret_access_key"
# Should return nothing!
# 5. Commit and push
git add .
git commit -m "feat: your feature description"
git push origin feature/your-feature-nameThis project is licensed under the MIT License - see the LICENSE file for details.
- AI4Bharat for the initiative
- AWS for Bedrock and infrastructure services
- Open-source community for amazing tools and libraries
- Documentation: Start with SETUP_SECRETS_GUIDE.md
- Issues: Please use GitHub Issues
- Security: For security vulnerabilities, please email security@zerotrust-ai4bharat.org
If you find this project useful, please consider giving it a star! β
Built with β€οΈ by the ZeroTRUST AI4Bharat Team
Last Updated: March 3, 2026