The research reveals that traditional service delivery systems are inadequate in predicting the use of e-government applications compared to artificial intelligence. Classic applications often fail to consider the differences in service delivery to citizens and recipients of electronic services, leading to a difference between classic applications and processing requests in complex service systems. There are currently no effective applications, methodologies, or methods to solve the problems of adopting artificial intelligence and linking it with e-government applications in Jordan. This paper aims to develop and study a way to combine artificial intelligence with classic e-government applications, using iris recognition as an example. The study focuses on e-government system applications that use iris biometrics to authenticate users accessing e-government systems. The theoretical and methodological basis of the paper is artificial intelligence using deep learning and convolutional neural networks to integrate this methodology into e-government services, improving citizens' access to services and determining the importance of recognizing people through iris for e-government applications as part of artificial intelligence (pp.25-32).