NIST releases an open-source privacy-preserving federated learning testbed with a curated soybean genomic dataset

NIST's National Cybersecurity Center of Excellence released the open-source Privacy-Preserving Federated Learning (PPFL) Testbed, software to train ML models centrally or across federated sites with or without differential privacy, plus a machine-learning-ready soybean genomic dataset of over 18,000 accessions from the USDA Soybean Germplasm Collection. Released 29 October 2026; genomics is the first use case.