The Open Source Initiative (OSI) has announced the launch of a new Open Source AI Fellowship, unveiled during UN Open Source Week. The program is designed to support OSI’s efforts to advance understanding, measure trends, engage in policy, and build consensus around what it means for an AI system to be labeled “Open Source.” Through the fellowship, OSI says it aims to act as a hub for Open Source AI, coordinating evidence, partnerships, and engagement on the key policy and governance questions shaping the field.
Central to the initiative is Gabriel Toscano, who was selected in May to serve as the Founding Fellow. According to OSI, he was chosen for his academic and professional achievements, his interdisciplinary acumen, and his demonstrated commitment to the organization’s mission. The two-year fellowship is grounded in a new partnership between OSI and Duke University’s Sanford School of Public Policy, an arrangement intended to foster principled, evidence-based, and stakeholder-driven engagement on the most pressing issues in Open Source AI.
The move builds on OSI’s earlier work defining the field. Version 1.0 of the Open Source AI Definition (OSAID) was released in October 2024 at All Things Open in Raleigh, North Carolina. Since then, OSI gathered feedback from its community and concluded that it needed a program broader than the definition itself, one capable of building research, policy, practice, and community around Open Source AI rather than resting on a single document.
The fellowship’s planned activities span several fronts. OSI intends to host multi-stakeholder thematic workshops examining the impact of the Open Source AI Definition and possible revisions, the role of open source in sovereign technology and the AI stack, evolving definitions and policy implications, and ways to expand access to AI as a digital public good. It also plans to develop a global research agenda, growing its collaboration with Open Forum Europe, the Open Knowledge Foundation, and the Digital Infrastructure Insights Fund to build the Open Technology Research Network established in November 2025, alongside ongoing policy and research analysis that tracks global AI governance against OSAID principles.
OSI is partnering with academic, civil society, and industry organizations to advance the freedom to use, study, share, and modify AI systems, positioning itself as a bridge between open source practice and AI policy. The organization credited Duke University along with launch sponsors Red Hat, AWS, Google, Automattic, and Mozilla for making the work possible, and said organizations interested in supporting the effort can reach OSI Executive Director Duane O’Brien at fellowship@opensource.org. Further details are available in the Open Source Initiative’s press release.