Highlights:s

  • The Open Model Initiative (OMI) was created in response to recent changes in licensing terms by developers of popular open-source models.
  • The OMI’s primary aim is to support the development of truly “open-source, capable, and ethical” generative AI models.

Recently, the Linux Foundation has launched a new initiative designed to foster the creation of high-quality, openly licensed artificial intelligence models. This initiative, which responds to the increasing need for genuinely open AI, aims to make these models freely accessible and to “drive creativity forward.”

The Open Model Initiative (OMI) was established in response to recent alterations in licensing terms by creators of prominent open-source models. Notably, Stability AI Ltd. recently updated the licensing for its widely-used image-generation model, Stable Diffusion 3 (SD3). Previously fully open and free, the model now requires a monthly fee and imposes new usage restrictions.

Stability AI faced criticism for its unclear licensing terms, but it’s not the only company that has imposed restrictions on software that was previously available for free.

The OMI aims to remove all obstacles to enterprise adoption by concentrating on training and developing AI models with “irrevocable open licenses without deletion clauses or recurring costs for access.”

The initiative, founded by organizations such as Invoke AI Inc., CivitAI Inc., and Comfy Org, has garnered support from community projects, including the Sentient Foundation and Wand Synthesis AI Inc. It has been recognized as an “official open-source foundation” under the Linux Foundation, with governance overseen by a community-based Steering Committee.

The OMI declares that its primary objective is to advance the creation of truly “open-source, capable, and ethical” generative AI models. To achieve this, it plans first to set up a governance framework and working groups to encourage collaborative community development and conduct a survey designed to collect feedback from the AI community on future model research and training.

The OMI’s plan includes establishing “shared standards” to improve AI model interoperability and metadata practices, as well as developing a transparent, open-source dataset for AI training accessible to all developers. Additionally, it aims to create an alpha test model for focused red-teaming and to release the initial version of this model, along with fine-tuning scripts, by the end of the year.

Linux Foundation Executive Director Jim Zemlin stated that the organization is dedicated to promoting open and collaborative development in the field of AI. He promised, “With the Open Model Initiative, we are taking a significant step toward making AI accessible and beneficial for everyone, building an environment where creativity and progress in AI can thrive without barriers.”

Kent Kiersey, Founder and CEO of Invoke, expressed his conviction that open-source is the most effective approach for AI and that it will benefit everyone. He said, “Through OMI’s commitment to open source development under the Linux Foundation, we believe the collective will deliver high-quality, competitive models with open licenses that push AI creativity forward, are free to use, and meet the needs of the community.”

The OMI welcomes anyone enthusiastic about open-source software and AI to get involved. They can join through the official GitHub repository and Discord channel.