More than three decades ago, Red Hat saw the potential of open source development and licensing to create better software and foster IT innovation. Thirty million lines of code later, Linux has not only evolved to become the most successful open source software but also maintains that position to this day. The commitment to open source principles continues, not only in the corporate business model but also as part of the work culture. The company believes these concepts have the same impact on artificial intelligence (AI) if done correctly, but the tech world is divided on what the “right way” would be.
AI, especially the large language models (LLMs) behind generative AI (gen AI), cannot be viewed the same way as open source software. Unlike software, AI models primarily consist of numerical parameter models that determine how a model processes inputs, as well as the connections it makes between various data points. Trained model parameters are the result of a lengthy process involving vast amounts of carefully prepared, mixed, and processed training data.
Although model parameters are not software, in some ways they serve a similar function to code. It’s easy to draw the analogy that data is the model’s source code or something very close to it. In open source, source code is commonly defined as the “preferred form” for making modifications to software. Training data alone does not fit this role, given its varying size and the complex pre-training process that results in a tenuous and indirect connection between any item of the data used in training and the trained parameters and the resulting behavior of the model.
Most of the improvements and enhancements in AI models happening now in the community do not involve access to or manipulation of the original training data. Instead, they result from modifications to model parameters or a fine-tuning process that can also serve to adjust the model’s performance. The freedom to make these model improvements requires that the parameters be released with all the permissions users receive under open source licenses.
Red Hat’s Vision for Open Source AI.
Red Hat believes the foundation of open source AI lies in open source licensed model parameters combined with open source software components. This is a starting point for open source AI, but not the final destination of the philosophy. Red Hat encourages the open source community, regulatory authorities, and the industry to continue striving for greater transparency and alignment with open source development principles when training and fine-tuning AI models.
This is Red Hat’s vision as a company, spanning an open source software ecosystem, to engage practically with open source AI. It is not an attempt at a formal definition, such as the one the Open Source Initiative (OSI) is developing with its Open Source AI Definition (OSAID). This is the corporation’s perspective on making open source AI feasible and accessible to the broadest set of communities, organizations, and vendors.
This perspective is put into practice through work with open source communities, highlighted by the InstructLab project, led by Red Hat, and the collaboration with IBM Research on the family of Granite open source licensed models. InstructLab significantly lowers barriers for non-data scientists to contribute to AI models. With InstructLab, domain experts from all sectors can add their skills and knowledge, both for internal use and to help build a shared, widely accessible open source AI model for upstream communities.
The Granite 3.0 model family addresses a wide range of AI use cases, from code generation to natural language processing for extracting insights from large datasets—all under a permissive open source license. We helped IBM Research bring the Granite code model family to the open source world and continue to support the model family, both from an open source perspective and as part of our Red Hat AI offering.
The impact of recent DeepSeek announcements shows how open source innovation can influence AI, both at the model level and beyond. There are naturally concerns about the Chinese platform’s approach, primarily that the model’s license does not explain how it was produced, highlighting the need for transparency. That said, the disruption mentioned reinforces Red Hat’s vision for the future of AI: an open future focused on smaller, optimized, and open models that can be customized for specific enterprise data use cases anywhere in the hybrid cloud.
Expanding AI Models Beyond Open Source
Red Hat’s work in the open source AI space extends far beyond InstructLab and the Granite model family, reaching into the tools and platforms needed to actually consume and productively use AI. The company has become very active in fostering tech projects and communities, including (but not limited to):
● RamaLama, an open source project aimed at simplifying the local management and deployment of AI models;
● TrustyAI, an open source toolkit for building more responsible AI workflows;
● Climatik, a project focused on making AI more sustainable in terms of energy consumption;
● Podman AI Lab, a developer toolkit focused on making experimentation with open source LLMs easier;
The recent announcement about Neural Magic expands the corporate vision for AI, making it possible for organizations to align smaller, optimized AI models, including open source licensed systems, with their data, wherever it resides in the hybrid cloud. IT organizations can then leverage the vLLM inference server to power the decisions and output of these models, helping to build an AI stack based on transparent and supported technologies.
For the corporation, open source AI lives and breathes in the hybrid cloud. The hybrid cloud provides the flexibility needed to choose the best environment for each AI workload, optimizing performance, cost, scale, and security requirements. Red Hat’s platforms, goals, and organization support these efforts, alongside industry partners, customers, and the open source community, as open source in artificial intelligence is propelled forward.
There is immense potential to expand this open collaboration in the AI space. Red Hat envisions a future that includes transparent work on models, as well as their training. Whether next week or next month (or even sooner, given AI’s rapid evolution), the company and the open community as a whole will continue to support and adopt efforts to democratize and open up the world of AI.
