The pathogen Alternaria porri (Ellis) Cifferi, is responsible for causing massive losses to the onion seeds crop by the formation of elliptical lesions on the flowering stalk, thus the plant topple over from there. On the basis of the present study the “seed-borne” nature of the pathogen was assessed. The fungus is responsible for causing Purple blotch disease in onion. The presence of the pathogen in seed was investigated with the help of various incubation methods. The best method found for the optimum recovery, i.e. ,53.10% in untreated as compared to 21.44% in pretreated seeds was the deep freeze method. All these methods of seed detection are extremely challenging because of time-taking experiments. A concept of advanced technologies such as Artificial Intelligence is highly appreciable for seed quality detection. A CNN model using advance AI-based systems is proposed for seed detection. With the help of this model early diagnosis of fungus is significant because it can minimize economic loss as well as unnecessary and injudicious use of fungicides. This paper reveals the use of Artificial Intelligence-based technology for quick diagnosis of infected seeds and survival of pathogen in seed.