Hugging Face and CMU Launch Project ProtoStar to Make Publicly Funded Science AI-Ready

Details

Project ProtoStar is a multi-institutional initiative led by Hugging Face, Carnegie Mellon University, and the Open Research Community Accelerator to make publicly funded scientific datasets, models, and code accessible in AI-ready formats.
The U.S. federal government invests approximately $190 billion annually in research and development, yet much of this data remains scattered in incompatible or legacy formats that hinder reuse by AI systems.
Hugging Face currently hosts over 17,000 public datasets in core scientific domains, with researchers sharing more than three times as many new science datasets in the first nine months of 2026 compared to the same period in 2025.
The inaugural open challenge, hosted by Carnegie Mellon University from November 2 to 13, 2026, tasks participants with using historical weather data from 210 U.S. cities to predict urban climate conditions and propose adaptation policies.
Founding partners include Purdue University, Georgia Institute of Technology, Arizona State University, The George Washington University, and the Astera Institute, all working through the Open Forum for AI.
Participating organizations formalize their involvement through a memorandum of understanding to support the release of suitable scientific artifacts on open infrastructure.

Key point