Announcing Project ProtoStar: Making Publicly Funded Science AI-Ready

Why it matters

Project ProtoStar is a new initiative aimed at making publicly funded scientific research—datasets, models, and code—genuinely reusable, especially for AI-driven work. Its founding partners include Carnegie Mellon University, Purdue University, Georgia Institute of Technology, Arizona State University, and The George Washington University, coordinated through the Open Forum for AI, alongside the Astera Institute, Hugging Face as infrastructure partner, and the Open Research Community Accelerator as community partner. The context is a gap between policy and practice: federal funders increasingly require public sharing of research outputs, but artifacts are often scattered, poorly documented, or stored in formats that neither humans nor AI systems can readily use. ProtoStar addresses this by encouraging institutions to release artifacts in AI-ready, machine-actionable formats on open infrastructure, formalized through memoranda of understanding, and by demonstrating value through open challenges. The first challenge, hosted by Carnegie Mellon University, focuses on climate risk intelligence. The initiative builds on prior efforts around data readiness and reflects accelerating growth in scientific dataset sharing on Hugging Face, suggesting real demand for reusable research resources.