- The latest release from Stability AI, Stable Diffusion 3 (SD3), is generating considerable controversy due to its licensing terms.
- SD3’s licensing terms are causing concern within the AI community regarding its impact on future development and collaboration.
- Kent Keirsey, CEO of InvokeAI, has highlighted key issues with the new licensing terms, indicating substantial implications for developers.
Discover how the new SD3 licensing terms could reshape the AI development landscape and why it’s causing upheaval in the community.
CivitAI Imposes Temporary Ban on SD3 Resources
The largest repository for Stable Diffusion resources, CivitAI, has announced a provisional ban on all resources related to Stable Diffusion 3 (SD3). The decision was taken to mitigate legal risks associated with the new licensing terms.
“Due to the ambiguity surrounding SD3’s licensing, we are temporarily banning all SD3-based models and related resources,” read the official statement. This move highlights growing concerns that the license grants Stability AI excessive control over model derivatives, sparking debates within the AI community.
Community’s Response to Licensing Terms
The new licensing terms of SD3 have raised alarms among developers, including the creator of Pony Diffusion, who has publicly criticized the lack of clear communication from Stability AI on the commercial use of SD3. Developers are facing limitations on image generation and stricter licensing policies, which hinder their ability to innovate freely.
Developers like InvokeAI have expressed frustration over the informal and non-binding responses from Stability AI. The lack of clarity has prompted some to continue using older models like SDXL, awaiting a more developer-friendly licensing model.
Implications for Open-Source AI Development
The restrictive licensing terms of SD3 pose significant challenges for the open-source AI community. Intellectual property concerns and the potential for perpetual fees are major issues that developers must navigate. The clause requiring destruction of derivative works if fees are not paid adds to the uncertainty, affecting trust and openness in the community.
This situation has prompted several enterprises to stick with SDXL, Stability AI’s prior model, until a more accommodating licensing structure is introduced. The controversy underscores a critical juncture for open-source AI and its future direction.
Conclusion
The arrival of Stable Diffusion 3 has created a legal and operational quagmire for the AI community. Developers are wary of the restrictive and potentially stifling licensing terms, leading some to revert to earlier models. The need for Stability AI to address these concerns clearly and promptly is crucial to maintain their developer base and support future innovation.
