OSAID 1.0 - The world's first open-source official AI standard version 1.0 released by OSI.

What is OSAID 1.0?

OSAID 1.0 (The Open Source AI Definition – 1.0) is an official standard released by the Open Source Initiative (OSI) to clarify the conditions for AI systems to be open source. More than 25 organizations participated in the development of OSAID 1.0, including Microsoft, Google, Amazon, Meta, Intel, Samsung, the Mozilla Foundation, the Linux Foundation, and the Apache Software Foundation. OSAID 1.0 requires AI systems to freely provide the necessary code, data, and parameters under OSI-recognized legal provisions, ensuring that users can use, research, modify, and share AI systems without restrictions. OSAID 1.0 particularly emphasizes the importance of data and code in machine learning systems, supports the exclusion of certain training data under legal restrictions, and promotes transparency and collaborative development of AI technology.

Specific definition of OSAID 1.0

• Free useOpen-source AI systems should allow users to use the system for any purpose without requesting permission.
• Free researchUsers can study how the system works and inspect the components.
• Free to modifyThe system can be modified, including changing its output, to meet any purpose.
• Free sharingUsers can share the system with others, whether or not it has been modified, for any purpose.

These freedoms apply to the entire system as well as to individual elements within it. The prerequisite is access to the preferred form for modifying the system. OSAID 1.0 specifically states that for machine learning systems, modifications to the preferred form must include the following elements:

• Data InformationProvide sufficiently detailed data so that a skilled person can build a substantially equivalent system using the same or similar data. This includes a complete description of all data used for training, including the source, scope, and characteristics of non-shareable data, as well as how to acquire and select data, labeling procedures, data processing, and filtering methods. It should also list all publicly available training data and how to obtain it, and all training data obtainable from third parties and their acquisition methods (including paid acquisition).
• codeProvide complete source code for training and running the system. The code should demonstrate a complete specification of how data is processed and filtered, and how training is performed. The code should be provided under an OSI-approved license.
• parameterProvide model parameters, such as weights or other configurations. Parameters should be provided in accordance with the terms approved by OSI.

For machine learning systems, an AI model consists of a model architecture, model parameters (including weights), and inference code to run the model. AI weights are a set of learned parameters that generate outputs based on given inputs. "Open source models" and "open source weights" must include the data and code used to derive these parameters.

OSAID 1.0 project address

• Project address:opensource.org/ai/open-source-ai-definition

Who is OSAID 1.0 suitable for?

• Open source communityProviding clear standards and guidance for open-source AI projects helps improve project quality and reliability, and attracts more developers to participate.
• enterpriseThis will affect companies' AI model release strategies, requiring adjustments to ensure compliance with OSAID standards, including technical, legal, and business-related changes.
• DevelopersIt provides more choices and flexibility, allowing developers to select and use open-source AI systems according to the OSAID standard, thus promoting innovation and customized development.
• Educational and research institutionsBased on open source, educational institutions and researchers can more easily access and use AI technology, promoting the development of AI education and research.
• Laws and policy makingIt provides a reference framework for policymakers, helping to formulate and adjust AI-related laws and policies, and ensuring that the legal system keeps pace with technological development.

OSAID 1.0's Industry Advantages

• ClarityOSAID 1.0 provides a clear set of definitions and standards to help developers, businesses, and users understand what true open-source AI is and how to ensure that AI projects comply with open-source principles.
• Promote innovationEncouraging openness and sharing, OSAID 1.0 helps foster innovation in the field of AI. Developers have the freedom to access and modify the code of open-source AI projects, driving the development of new technologies and applications.
• Enhance trustAdherence to the OSAID 1.0 standard increases user trust in AI systems. Open source implies transparency, allowing users to verify how the system works, which helps build trust in AI technology.
• Collaboration opportunitiesOpen-source AI projects can attract the participation of developers worldwide. The OSAID 1.0 project provides a common framework that makes collaboration easier and more efficient.

Scope of OSAID 1.0

• Impact on the development and use of AI modelsOSAID stipulates that open-source AI models must provide sufficient information so that anyone can "substantially" reconstruct the model, including the source, processing, and acquisition of the training data.
• Promoting AI transparency and opennessOSAID emphasized the importance of transparency, requiring open-source AI models to disclose the complete source code used to train and run AI, as well as the model parameters.
• Promote the healthy development of AI technologyIt provides clear guidance and standards for the development and application of open-source AI.
• Impact on existing AI modelsMany AI models that claim to be "open source" do not actually conform to OSAID standards, which may have an impact on the development and use of these models.
• Legal and copyright issuesThis will have an impact on the legal status and copyright issues of AI models, especially regarding whether the models can be protected by copyright.