Action #5
In Progress

Make AI visible in works (BadgeAI.org)

Develop a standardized AI transparency badge system inspired by Creative Commons, allowing creators to precisely indicate AI involvement in their works (Audio, Visual, Text) with 5 contribution levels

Difficulty:Hard
Category:technical
Impact:international

Badge AI - Transparency System (In Development)


Current Project Status


🚀 Active website: badgeAI.org


Phase 1 - Self-declaration:Functional

  • Web interface allowing creators to generate AI badges
  • 5 transparency categories (0 = Human only, 4 = AI only)
  • 3 media types: Audio (S), Visual (V), Text (T)
  • System inspired by Creative Commons

  • Phase 2 - Automatic detection: 🔄 In development (very difficult)

  • Integration into creation software (Photoshop, ChatGPT, etc.)
  • Automatic metadata with cryptographic signature
  • Standardized API for platforms
  • Real-time AI usage detection

  • The Badge AI System


    5 Contribution Levels

  • 0: Human Only
  • 1: Human with AI Assistance
  • 2: AI Collaboration (balanced)
  • 3: AI Directed (human prompts)
  • 4: AI Only

  • 3 Media Types

  • S-AI-0-4: Audio (Audio, music, voice, effects)
  • V-AI-0-4: Visual (Images, videos, animations)
  • T-AI-0-4: Text (Written content, code, data)

  • Technical Vision

  • Metadata embedded in files (EXIF, ID3, HTML meta)
  • Cross-platform interoperability
  • Complete history of AI modifications
  • Anti-tampering cryptographic signatures

  • Major Technical Challenges


    ✅ **Easy - Already implemented**

  • Web badge generation interface
  • Creator self-declaration
  • Standardized notation system

  • 🔥 **Very Difficult - In progress**

  • Automatic detection in software
  • Native integration in Adobe, Google, OpenAI
  • Persistent cross-platform metadata
  • Badge tampering prevention

  • Potential Impact


  • Universal transparency: Every AI content clearly identified
  • Public trust: Clear human/AI distinction
  • Creator protection: Human work valorization
  • International standard: Tech industry adoption

  • ---


    Project developed by Philippe Bourque (Cérebrum) - Contribute on GitHub