Abstract
Plant viral infections continue to pose a significant and ongoing threat to global food security, especially in the context of climatic instability and intensive agricultural practices. The CRISPR/Cas system has emerged as a powerful tool for developing virus-resistant crops by enabling precise modifications to viral genomes or plant susceptibility factors. Nonetheless, the efficacy and dependability of CRISPR-based antiviral approaches are limited by challenges in guide RNA design, off-target effects, insufficiently annotated datasets, and the intricate biological dynamics of plant-virus interactions. This paper summarizes the latest advancements in the incorporation of artificial intelligence (AI) methodologies, including machine learning and deep learning algorithms, into the CRISPR design and optimization framework. It examines how convolutional and recurrent neural networks, transformer architectures, and generative models like AlphaFold2, RoseTTAFold, and ESMFold can be used to predict protein structures, score sgRNAs, and model host-virus interactions. AI-enhanced methods have been proven to improve target specificity, Cas protein performance, and in silico validation. This paper aims to establish a foundation for next-generation genome editing strategies against plant viruses and promote the adoption of AI-powered CRISPR technologies in sustainable agriculture.