The landscape of artificial intelligence is evolving rapidly, with breakthroughs in large language models, privacy technologies, and governance frameworks reshaping how researchers, developers, and policymakers think about responsible AI. But as AI systems become more powerful and more widely deployed, the question of how these technologies should be developed — and for whom — is gaining urgency.
Enter the Open Forum for AI (OFAI): an international initiative launched in 2024 and led by Carnegie Mellon University that brings together academic institutions, non-profit organizations, and domain experts to guide the future of AI toward human-centered, transparent, and ethical outcomes.
OFAI recognizes that AI must be shaped not just by engineers and corporations, but by a diverse ecosystem that includes civil society, governance experts, educators, and open source advocates.
This initiative’s mission couldn’t be more timely. As AI systems touch every part of society, from healthcare and scientific research to government services and education, it’s essential that AI development is grounded in values like fairness, accountability, and inclusivity. OFAI aims to advance these principles by fostering collaboration, informing policy discussions, and building platforms that empower communities to contribute to AI design and governance.
In a significant step forward for the movement toward responsible AI, OpenMined — the nonprofit community and foundation co-created and led by Andrew Trask — has joined OFAI as part of its Data Working Group. OpenMined’s mission is to make privacy-preserving technologies accessible and practical for researchers and developers around the world. The organization designs open-source tools and protocols that make it possible to train and evaluate AI on sensitive data without compromising privacy, a key requirement for building trustworthy AI systems.
For years, Andrew Trask and the broader OpenMined community have championed the idea that privacy and openness aren’t mutually exclusive — and that by combining novel techniques like federated learning, differential privacy, encrypted computation, and other privacy-enhancing technologies, we can build AI systems that are both powerful and responsible. These tools allow organizations to collaborate on AI workflows without exposing raw data, protecting individuals’ privacy while unlocking valuable insights from datasets that were previously inaccessible.
OpenMined’s participation in OFAI signals a deeper alignment between the technical infrastructure needed for responsible AI and the governance frameworks guiding its evolution. As part of the Data Working Group, OpenMined will contribute expertise in how data can be governed, shared, and stewarded to support transparency, equity, and accountability. This involves mapping key stakeholders in open AI infrastructure, crafting policy and governance frameworks for responsible data use, and promoting practices that ensure AI systems reflect human values.
According to Andrew, “To foster public trust, AI has to be built on data practices people can understand and rely on.” His organization’s involvement in OFAI creates an essential bridge between communities building privacy-focused AI tools and the broader coalition working to shape AI’s ethical and social contours.
Bottom line:
As AI continues to integrate into every aspect of life, initiatives like OFAI — strengthened by contributors like OpenMined — are vital to ensuring that AI innovation does not leave behind the principles of human dignity, transparency, and societal benefit. Through collaboration and shared expertise, this partnership exemplifies how diverse stakeholders can work together to build an AI ecosystem that truly serves the public good.
Invite Andrew to your next event. Find out more here.