We’re excited to announce that OpenMined has joined the AI Alliance, a global community of leading technology developers, researchers, and adopters collaborating to advance open, safe, and responsible AI.
About the AI Alliance
Launched in December 2023 by IBM and Meta, the AI Alliance now brings together over 100 organizations from established firms like Intel, AMD, Dell, Oracle, and Sony, to leading AI startups like Hugging Face, Stability AI, and Together AI, to world-class academic institutions including Cornell, ETH Zurich, UC Berkeley, and the University of Tokyo.
What drew us to this community is its clear recognition that data is foundational to AI. The Alliance has organized working groups around Foundation Models and Datasets, Trust and Safety, and other critical areas where OpenMined is excited to offer its expertise.
Why Data Governance Matters for Open AI
The AI Alliance is committed to responsibly advancing the ecosystem of open foundation models, but without good datasets, model training and tuning would be impossible. That’s why the Alliance is promoting the development of open datasets with clear governance and provenance controls so they can be used without concerns about legal and other risks.
This mission resonates deeply with us. At OpenMined, we’ve spent years building the infrastructure to enable exactly this kind of responsible data access, but from a complementary angle. While the Alliance focuses on curating openly-licensed datasets with transparent governance, there’s a parallel challenge: vast amounts of valuable data can never be made fully open due to legitimate privacy, security, and confidentiality requirements. Medical records, proprietary research, financial data, and user behavior logs hold enormous potential to advance AI, but sharing them directly poses unacceptable risks. Through privacy-enhancing technologies such as federated learning, differential privacy, and encrypted computation, we enable AI development on sensitive data without requiring the data to leave its secure environment. Researchers can train models, validate findings, and extract insights, all while the underlying data remains protected.
What’s Next
We’re looking forward to engaging with Alliance members across the Foundation Models and Datasets working group, the Trust and Safety initiatives, and beyond. This partnership represents another step in our mission to build the infrastructure for responsible AI development, where data can be used to advance science and benefit society while respecting the privacy of the individuals and organizations who create it.
