CMU OSAID Statement

The following Carnegie Mellon University (CMU) faculty express support for the Open Source Initiative (OSI) Open Source AI definition (OSAID). We believe the OSAID reflects a balanced accounting for legal considerations, particularly given diverse copyright provisions globally, and practical constraints with large datasets. It also recognizes the data-mediated, computational and engineering uncertainty of AI systems. There remain unresolved research questions into the nature of how AI systems work.

Additionally, while software remains an important component, AI is fundamentally more complex. It is worthwhile to consider how the freedoms of open source software may transfer to open source AI, but the specific software development frameworks and practices do not map seamlessly to AI, as reflected within the OSAID.

The Mozilla Foundation noted the importance of OSAID in its statement of support, calling it a “critical junction in the evolution of the internet” and “about shaping the technology and its impact on society.” We recognize that OSAID can become part of a broader framework for defining openness in AI, which represents an important component of a public AI ecosystem.

Given the potential impact of the OSAID, we applaud the OSI for adopting a participatory co-design process which, in the aggregate, was inclusive and participatory. As noted on its website, OSI held a series of convenings over two years with a diverse range of participants and stakeholders from the open source software community, spanning the private, public and non-profit sectors.

Noting the dynamic nature of AI, the unresolved research questions, and its potential influence for policy or regulatory development, we call upon the OSI to continue evolving the definition through a co-design process, which specifically reaches out to the university sector. The OSI is a partner in the CMU led Open Forum for AI (OFAI), which offers a natural pathway for such engagement.

We welcome the opportunity to work with OSI toward evolving the OSAID definition toward its next versions.

Signed by:

Ganesh Mani, Distinguished Service Faculty and Director of Collaborative AI (Tepper) & Adjunct Faculty (S3D/SCS)
Hoda Heidari, K&L Gates Career Development Assistant Professor in Ethics and Computational Technologies, School of Computer Science, Carnegie Mellon University
Emily Bongiovanni, Open Knowledge Librarian, Carnegie Mellon University Libraries
Melanie Gainey, STEM Librarian, Director of Open Science Program, Carnegie Mellon University Libraries
Keith Webster, Helen & Henry Posner Jr Dean of University Libraries and Director of Emerging and Integrative Media Initiatives, Carnegie Mellon University
Di Yoong, Social Sciences Librarian, Carnegie Mellon University Chasz Griego, STEM Librarian, Carnegie Mellon University Libraries Sam Lemley, Curator of Special Collections, Carnegie Mellon University Libraries
Prasad Chalasani, Adjunct Professor of AI (Tepper), CoFounder of Langroid
Haoyong Lan, STEM Librarian, Carnegie Mellon University
Reya Saliba, Instruction & Outreach Librarian, Carnegie Mellon University in Qatar
Sarah Young, Social Sciences Librarian and Director of the Evidence Synthesis Program, Carnegie Mellon University Libraries, Pittsburgh, PA
Anand Rao, Distinguished Service Professor of Applied Data Science and AI, Heinz College of Information Systems and Public Policy, Carnegie Mellon University
Rachel C. Kurchin, Assistant Research Professor of Materials Science and Engineering, Carnegie Mellon University
Huajin Wang, STEM Librarian, Carnegie Mellon University
Assane Gueye, Associate Teaching Professor, Carnegie Mellon University Africa
Vincent Sha, Associate Dean of IT & Operations, Dietrich College, Technical Lead - Open Forum for AI, Carnegie Mellon University
Nicky Agate, Associate Dean for Academic Engagement, Carnegie Mellon University