In this research paper, a new method for identifying strawberry leaf diseases using federated learning along with Convolutional Neural Networks (CNN) is proposed. The study addresses five types of strawberry leaf disorders, involving five different clients collectively to devise a strong and datapreserving diagnostic method. A central component of the approach includes training local models on client-specific data and aggregating them globally by employing federated averaging techniques. The local data analysis results from the five clients i.e hg_1, hg_2 to upto hg_5 were promising as it informed the model’s effectiveness in disease detection and classification. The performance metrics for each client were as follows: 90.54%, 90.15%, 90.5%, 89.06 and 96.25%, macro average of hg_1 to hg_5. The local data results effectively converted into a global model by the federated averaging process promoted enhanced overall performance while preserving privacy. The consolidation of local data to global was another factor, which ensured that each client provided the same amount information towards building a global model through federated averaging. In summary, this research provides evidence that federated learning with CNNs has great potential for use in the agricultural industry and more specifically to diagnose diseases on leaf surfaces of strawberries precisely and efficiently. The approach presents a scalable, decentralized and privacy-conscious solution that allows AI to be applied in agriculture