Open Commercial Source: A Framework for a New Era

This is part two of a two-part series on the future of Open Source in the era of Generative AI. To view part one, click here.

Introducing Open Commercial Source License

It's becoming increasingly clear that we need a new model—one that acknowledges the unique realities of generative AI while preserving the innovation-driving principles of open source.


At Bria, we propose what we call an "Open Commercial Source License" that redefines the four freedoms for the generative AI era:

1. Run: Free for non-commercial use, pay for commercial use, and built on licensed data with transparency
2. Study: Documentation is fully available, enabling understanding of the system
3. Modify: Source code for both inference and training is available, allowing customization
4. Redistribute: Limited redistribution rights that respect data ownership through a patented managed attribution model

A modified definition of the Four Freedoms for the Gen AI era.

This approach strikes a balance between openness and sustainability. It ensures that:

• Developers can access the complete technical stack for innovation
• Data owners are respected and fairly compensated through managed attribution that tracks works influence on outputs
• Commercial viability is maintained to fund ongoing development
• Safe commercial use is enabled with clear rights and responsibilities

Implementing the Vision: Bria's Approach

At Bria, we've built our visual generative AI platform on these principles. Our foundation models are trained exclusively on licensed data, with an attribution engine that rewards data owners based on their content's impact on generated outputs.

The Bria Platform uses a patented attribution engine to reward artists and data owners
by their contribution to generated outputs. 

This approach ensures that:

1. Data owners receive fair compensation based on usage
2. Users can develop without copyright or privacy infringement concerns
3. The ecosystem becomes sustainable for all participants

Our implementation provides developers with:

• Source code availability for all models and weights
• Complete building blocks for image generation, editing, and more
• Flexible deployment options across cloud platforms
• Enterprise-grade compliance with full indemnity

APIs, models, and pipelines enable complete development control and flexibility..

This enables innovation while respecting the complex realities of data ownership, compute requirements, and responsible development.

The Path Forward: A Call for New Models

Generative AI presents unprecedented challenges to traditional open source concepts, but it doesn't mean we must abandon the spirit of innovation and collaboration that has driven technological progress.

What's needed now is a thoughtful evolution of licensing models that:

1. Respect data ownership, privacy, and copyright
2. Account for the significant compute costs of model development
3. Address the increased complexity of these technologies
4. Enable both innovation and commercial viability

The Open Commercial Source License model represents one approach, but we need more voices in this conversation. The frameworks we develop today will shape how AI evolves over the coming decades.

We find ourselves at a pivotal moment—similar to when Richard Stallman encountered that locked-down printer. The decisions we make now about openness, access, and sustainability will determine whether generative AI becomes a technology that empowers broad innovation or one that remains confined to a handful of powerful gatekeepers.

We invite you to join this conversation and help shape a future where generative AI development is both open and sustainable—where commercial viability and innovation coexist rather than conflict.

Start Building Now
Connect with Us