The International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), with the Leibniz Institute of Plant Genetics and Crop Plant Research (IPK), Germany, and the University of Queensland, Australia, introduced a scientific framework combining artificial intelligence with pangenetics to reshape crop breeding. Published in Molecular Plant, the research proposes integrating biology-driven crop genetics with advanced AI prediction models to accelerate development of climate-resilient, high-yielding and nutritionally improved crop varieties. The study addresses the challenge of balancing biological understanding with predictive analytics, noting conventional genomic prediction models often fail to capture biological interactions governing multiple characteristics, while functional genomics provides deeper insight but is validated across a narrower range of genetic materials and environmental conditions. The framework bridges these approaches by combining functional genomics, genetic diversity preserved in genebanks and AI into a unified breeding strategy. A key element is pangenetics, which links genetic variation directly with functional trait biology. The researchers suggest incorporating biologically informed genomic features into advanced AI models, including deep learning systems, could enable breeders to identify superior breeding lines more efficiently. The framework supports broader AI application throughout crop improvement pipelines, and combining AI with large-scale genomic and phenotypic datasets has the potential to shorten breeding cycles while improving precision. The research builds on ICRISAT's ongoing efforts to modernize crop improvement through digital technologies, including speed breeding and predictive breeding.