artshield

ArtShield

Democratizing Defense in the Age of AI

A local-first, privacy-focused tool for independent artists to protect their work from non-consensual AI training.

View Web Demo • Getting Started • Features

---

🛡️ The Mission

ArtShield is the “Signal” of image protection. In a landscape where creative work is scraped without permission, we provide a “One-Click Shield” that restores agency to the artist.

We believe:

1. Consent is Non-Negotiable: Artists should decide if their work updates a model.
2. Privacy is Paramount: Your art never leaves your machine. Processing happens locally on your GPU/CPU.
3. Access is Equity: Protection shouldn’t require a Computer Science degree.

✨ Features

• The Cloak (Mist v2): Integrated adversarial protection that disrupts style mimicry in AI models.
• The Tag (Metadata): Injects “Do Not Train” signals (NoAI tags, Copyright assertions) directly into file headers.
• The Stamp (Watermark): Optional, elegant visual watermarking for social media display.
• Local Execution: Powered by a Python backend that runs entirely on your device (Windows/Mac).
• Web Demo Mode: A browser-based “Mock Mode” to demonstrate the UI/UX without installing the desktop app.

🏗️ Technical Architecture

ArtShield uses a Hybrid Architecture to combine raw Python power with a modern, responsive UI.

• Frontend: React + Vite + Tailwind CSS
  • Why? For a premium, “Apple-like” aesthetic and responsiveness.
• Backend: FastAPI + PyTorch + Pillow
  • Why? To leverage the mist_v2 research code and efficient image processing.
• Packaging: PyInstaller
  • Why? To bundle everything into a single, double-click executable (.exe / .app).

🚀 Getting Started

Prerequisites

• Node.js 20+
• Python 3.10+
• (Optional) NVIDIA GPU with CUDA for faster processing.

Installation

1. Clone the Repository

   git clone https://github.com/jwlehane/artshield.git
   cd artshield
   

2. Setup Frontend

   cd frontend
   npm install
   

3. Setup Backend

   cd ../backend
   python -m venv venv
   # Windows:
   .\venv\Scripts\activate
   # Mac/Linux:
   source venv/bin/activate
       
   pip install -r requirements.txt
   

Running Locally

Option A: Full App (Backend + Frontend)

1. Start the API Server:

   # Terminal 1 (from root)
   python backend/main.py
   

2. Start the UI:

   # Terminal 2 (from root)
   cd frontend
   npm run dev
   

   Open http://localhost:5173. The UI will connect to the local Python API.

Option B: Mock Mode (Frontend Only) Run the UI with simulated API responses (perfect for UI dev or demos).

cd frontend
VITE_MOCK_API=true npm run dev

📦 Deployment & CI/CD

• Web Demo: Use the GitHub Actions workflow deploy.yml. It automatically deploys the frontend in “Mock Mode” to GitHub Pages on every push.
• Executable: (Coming Soon) The same workflow will run PyInstaller on Windows/Mac runners to generate downloadable artifacts.

🗺️ Roadmap

Phase	Status	Goal
I. MVP	🟡 In Progress	Basic Drag & Drop UI, Mist Integration, Local Build System.
II. Connected	⚪ Planned	Auto-updates for protection weights, “Verify My Shield” tool.
III. Collective	⚪ Planned	Decentralized Opt-Out Registry & Artist Identity Verification.

🤝 Contributing & Ethics

We encourage contributions from developers, researchers, and artists. Note: We strictly adhere to an ethical code—we do not build tools that facilitate non-consensual scraping or removing these protections.

---

Built with ❤️ by the DigitalHorizons Team.