What's Open Source AI?

Following the same idea behind Open Source Software,
an Open Source AI is a system made available under terms that grant users the freedoms to:

Use the system for any purpose and
without having to ask for permission.

Study how the system works and
understand how its results were created.

Modify the system for any purpose,
including to change its output.

Share the system for others to use with
or without modifications, for any purpose.

Precondition to exercise these freedoms is to have access to
the preferred form to make modifications to the system, and to the means to use it.

Benefits of Open Source AI

Transparency & Safety

Open Source AI provides information essential for auditing systems and to mitigate bias, ensures accountability and transparency of data sources, and accelerates AI safety research.

Competition & Polyculture

Open Source AI makes more models available, spurs innovation and quality due to increased competition and tackles AI monoculture by providing more stakeholders access to foundational technology.

Diverse Applications

Open Source AI gives developers access to resources crucial for developing context- specific, localized applications that are representative of cultural and linguistic diversity and allow for model aligned with different value systems.
OSAID Paris Workshop

Why Open Source AI needs a definition?

Open Source Frontier

The traditional view of Open Source code and licenses when applied to AI components are not sufficient to guarantee the freedoms to use, study, share and modify the systems.

Informing Regulators

Government regulations have begun in Europe, the United States, and elsewhere. Communities need a common understanding to educate policy makers.

Combat Openwashing

Companies are calling AI systems “open source” even though their licenses contain restrictions that go against the accepted principles and freedoms of open source.

Who's behind the Open Source AI Definition

View All Endorsers

Overall process

20+

Supporting Organizations

100+

Supporting Individuals

50+

Co-designers

13

Systems reviewed

Representation in the co-design process

27

Nationalities

42%

People of Color

33%

Global South

31%

Femme, Trans, & Nonbinary

Co-design

2023 - 2024

In 2023, we started the co-design process hosting several online and in-person activities around the world.



Research

2022 - 2023

Alongside AI experts from various fields we produced a podcast, panels, and webinars.

Endorsements

2024 - 2025

                                   Late 2024 into 2025, the OSI is gathering endorsements from various individuals and organizations, including Mozilla, Suse, Eleuther AI, Ai2, Eclipse Foundation, and the OpenInfra Foundation, among many others.

Which AI systems comply with the OSAID 1.0?

As part of our validation and testing of the OSAID, the volunteers checked whether the Definition could be used to evaluate if AI systems provided the freedoms expected. The list of models that passed the Validation phase are: Pythia (Eleuther AI), OLMo (AI2), Amber and CrystalCoder (LLM360) and T5 (Google). There are a couple of others that were analyzed and would probably pass if they changed their licenses/legal terms: BLOOM (BigScience), Starcoder2 (BigCode), Falcon (TII). Those that have been analyzed and don't pass because they lack required components and/or their legal agreements are incompatible with the Open Source principles: Llama2 (Meta), Grok (X/Twitter), Phi-2 (Microsoft), Mixtral (Mistral). These results should be seen as part of the definitional process, a learning moment, they're not certifications of any kind. OSI will continue to validate only legal documents, and will not validate or review individual AI systems, just as it does not validate or review software projects.

Co-design process

The OSAID co-design process is open to everyone interested in collaborating.

How to participate

There are many ways to get involved:

  • Join the working groups: be part of a team to evaluate various models against the OSAID.
  • Join the forum: support and comment on the releases, record your approval or concerns to new and existing threads.
  • Comment on version 1.0: provide feedback on the latest version of the document directly.
  • Follow the weekly recaps: subscribe to our newsletter and blog to be kept up-to-date.
  • Watch the town hall recordings to learn more about the process.
  • Join the workshops and scheduled conferences: meet the OSI and other participants at in person events around the world.
  • Endorse the Open Source AI Definition: have your organization appended to list of supporters of version 1.0.

Open Source AI Definition governance

Governance for the Open Source AI Definition is provided by the OSI Board of Directors. The OSI board members have expertise in business, legal, and open source software development, as well as experience across a range of commercial, public sector, and non-profit organizations. Formal progress reports including achievements, budget updates, and next steps are provided monthly by the Program Lead for advice and guidance as part of regular Board business. Additionally, informal updates on the outcomes of key meetings and milestones are provided via email to the Board as required.

Statements of support

"LLM360 finds that OSI’s Open Source AI definition is a meaningful, reasonable, and holistic standard which will have positive reverberations throughout the community. The definition clarifies the unique challenges surrounding open source AI including the expectations for disseminating code, data, and accessibility requirements. This definition propels the open source ecosystem and aligns with LLM360’s mission for community owned AI. Our team is thrilled and excited to fully support OSI’s efforts on advancing the Open Source AI definition."

Hector Zhengzhong Liu, LLM360

"Coming up with the proper open-source definition is challenging given restrictions on data, but I'm glad to see that the OSI v1.0 definition requires at least that the complete code for data processing (the primary driver of model quality) be open-source. The devil is in the details, so I'm sure we'll have more to say once we have concrete examples of people trying to apply this Definition to their models."

Percy Liang, Director of Center for Research on Foundation Models, Stanford University

“Facilitating an Open ecosystem is an important part of our approach at Intel. An open approach to AI can foster greater collaboration across the community, drive innovation and enhance transparency. We applaud OSI’s efforts to expand their definition to include AI models and datasets. OSI’s creation of a first revision of the definition, can help industry continue to evolve and iterate.”

Arun Gupta, Vice President and General Manager, Open Ecosystem, Intel

"We welcome OSI's stewardship of the complex process of defining open source AI. The Digital Public Goods Alliance secretariat will build on this foundational work as we update the DPG Standard as it relates to AI as a category of DPGs"

Liv Marte Kristiansen Nordhaug, CEO of the Digital Public Goods Alliance

"Transparency is at the core of EleutherAI’s non-profit mission. The Open Source AI Definition is a necessary step towards promoting the benefits of open source principles in the field of AI. We believe that this definition supports the needs of independent machine learning researchers and promotes greater transparency among the largest AI developers."

Aviya Skowron, Head of Policy and Ethics at Eleuther AI

"The Common Crawl Foundation fully supports the Open Source AI Definition as a crucial step in setting clear standards for open and transparent AI development.  This definition will help ensure AI develops responsibly, staying open and accessible to everyone."

Thom Vaughan, Principal Technologist, Common Crawl Foundation

"Transparency is at the core of EleutherAI’s non-profit mission. The Open Source AI Definition is a necessary step towards promoting the benefits of open source principles in the field of AI. We believe that this definition supports the needs of independent machine learning researchers and promotes greater transparency among the largest AI developers."

Stella Biderman, AI and NLP Researcher, EleutherAI

"SUSE applauds the progress made by the Open Source Initiative and its Open Source AI Definition. The efforts are culminating in a very thorough definition, which is important for the quickly evolving AI landscape and the role of open source within it. We commend the process OSI is utilizing to arrive at the definition and the adherence to the open source methodologies. Clarity and consensus drive collaboration, and we believe this definition will drive open source AI forward."

Alan Clark, Office Of The CTO, SUSE

"I endorse! We need common vocabulary to define what is open is what isn't. This is a solid framework that doesn't give a blank check to those who are lightly claiming to be providing open source AI (even if they desperately wish to be qualified as such), and reversely, the framework is open to initiatives that introduce gradients of open source on the various components that make an AI system, and recognizes efforts in opening-up all or some of the components. After all, "AI" is a derivative of software, complete with data, code and artefacts. There is no reason a derivative system should be classified under the foundational definition of "open source" and at the same time, AI systems are becoming so powerful at capturing intelligence away from humans that we need to qualify their degree of openness. Hats off to all involved for producing such an important piece of work."

Yann Lechelle, Co-founder CEO @ :probabl.

"This effort you and OSI team have been driving is really important and I’m a believer that time is becoming of the essence. Inevitably it will need to evolve but putting a stamp on it soon is important. We have to define what open source means in the context of AI models in order to preserve the permissionless innovation aspect that created so much value with open source software licenses. The definition is both pragmatic and challenging, and is an excellent first step in a fast-moving area."

Mark Collier, Chief Operating Officer at OpenStack Foundation

"The codesign process allowed me to see first hand the thought process of people all over the world about what is open source AI. It may never be possible for all the people to agree on the definition. But It is a wonderful start and I think everyone will agree that the open discussions, seminars, townhall meetings, follow up surveys, emails are all very effective and "democratic" :-)"

Victor Lu, Independent Consultant

"Open Source Group Japan commends OSI for its leadership in navigating the complex process of defining Open Source AI, and we fully support the Open Source AI Definition (OSAID) as a key standard for open and transparent AI systems. The field of AI is evolving rapidly, and the need for a clear and consistent definition of Open Source AI has never been more critical. OSI's OSAID marks a crucial milestone toward a future where collaboration and openness are the norms in AI development. We anticipate that this will drive innovation, transparency, and the ethical development of AI systems."

Shuji Sado, Chairman, Open Source Group Japan

"Open Source generative AI models are one of the keys to the advancement of the field. By enabling a community of developers and researchers to collaborate and evolve these models in a responsible way, we can greatly benefit a wide range of applications."

Oscar Mullin, VP of Technology - Cloud Services, Data & AI at MercadoLibre

Supported by

brand_start brand_start brand_start brand_start

OSI’s efforts wouldn’t be possible without the support of our sponsors and thousands of individual members.
Become a sponsor or join us today!