CCBL AI Policy

Overview

CCBL includes a built-in AI policy that protects software from being used to train AI systems without explicit permission. This addresses growing concerns about AI companies using open source code to train models without compensating or even acknowledging the original creators.

What is Prohibited

Using CCBL-licensed software to train commercial AI models without permission, including but not limited to:

• Large Language Models (LLMs): Training models like GPT, Claude, or Llama
• Generative AI: Training image, video, audio, or code generation models
• Machine Learning Models: Training classification, prediction, or recommendation systems
• Neural Networks: Any neural network training using the software or its code

Examples of Prohibited Use

1. Scraping code from a CCBL-licensed repository to train a code completion AI
2. Using the software's output to train an AI system
3. Analyzing the software's behavior to train an AI to replicate that behavior
4. Including the software in a dataset used for AI training

What is Allowed

✅ Using the Software as a User

End users can use CCBL-licensed software normally. The AI policy does not restrict normal usage.

✅ Using AI-Powered Tools

You can use AI-powered development tools (like GitHub Copilot, ChatGPT, etc.) to assist with:

• Writing code
• Debugging
• Documentation
• Testing

The AI policy restricts training AI systems, not using AI tools.

✅ Contributing Improvements

Contributors can submit improvements, bug fixes, and new features. Contributors receive a permanent commercial license as a thank you.

✅ Academic Research

Academic research about the software is allowed and encouraged.

✅ Non-Commercial AI Training

Personal research, academic study, or non-commercial AI training is permitted. However, if a model produced through non-commercial training is later used commercially, that commercial use remains subject to the license's commercial-use conditions (Section 3).

Why This Matters

The Problem

AI companies have been scraping open source repositories to train their models without:

• Asking permission from creators
• Compensating contributors
• Acknowledging the source of their training data

The Impact

• Devalues contributions: Contributors' work is used to create proprietary AI systems
• Reduces incentives: Less motivation to contribute to open source
• Concentrates power: AI benefits flow to large tech companies, not the community

CCBL's Solution

By requiring explicit permission for AI training, CCBL:

• Protects creators: Ensures their work isn't used without consent
• Enables negotiation: Allows creators to negotiate terms or compensation
• Promotes fairness: Creates a more equitable relationship between AI companies and open source

How to Request Permission

If you want to use CCBL-licensed software for AI training:

1. Contact the copyright holder listed in the LICENSE file
2. Explain your use case and how the software will be used
3. Negotiate terms if applicable
4. Get written permission before proceeding

Enforcement

Violations of the AI policy may result in:

• License termination: Your right to use the software ends
• Legal action: Potential copyright infringement claims
• Public notice: Violations may be publicly documented

FAQ

For common questions about the AI policy — including whether you can use AI tools with CCBL software, how to request AI-training permission, and how CCBL differs from other licenses — see the AI Policy section of the FAQ.