AI-driven tri-typing in agriculture: Current advances, future frontiers

Sustainable crop improvement is urgently needed to ensure global food security, particularly for developing and densely populated countries. The integration of artificial intelligence (AI) and machine learning (ML) into crop science tri typing is reshaping the conventional agriculture practices into an era of high-throughput phenotyping (HTPP) data-driven modern agriculture. AI tools accelerate data generation, mining, imputation, storage, transfer, and optimal decision-making within agricultural systems. AI tools are paving the way for modern plant breeding strategies by uncovering genetic variability and bridging the genotype-to-phenotype (G2P) gap, thus enabling the future of predictive breeding. Plant genetic gains or phenotype (P), by and large, depend on the genotype (G), environment (E), and their interaction (GEI). This review will provide a comprehensive overview of the historical background, current status, and prospects for integrating AI and ML tools in agricultural tri-typing, encompassing genotyping, phenotyping, and envirotyping. We explore AI-driven tools for genome analysis, HTPP platforms, and environmental data integration, emphasizing how these technologies overcome persistent bottlenecks in predictive breeding. Furthermore, this review will offer the reader key insight into modern trends, including the paradigm shift in phenomics patent filings, global distribution of HTP phenomics facilities, the publications volume and related research over the last two decades, and individual institutions currently leading or prospectively will lead the world in plant phenomics. Similar to plant phenotyping, we also try to address the integration and application of AI/ML algorithms in plant genotyping and envirotyping.

The online version contains supplementary material available at 10.1007/s11032-026-01705-1.