Rapid, appropriate, and reliable diagnosis is paramount for selecting a suitable therapeutic intervention for the clinical management of COVID-19. Several serological and molecular diagnostic methods are available, however, biosensor-based diagnosis has been employed in the diagnosis of viral diseases including COVID-19 due to its high specificity, sensitivity, expeditiousness, low cost, and capability to detect the analyte even at low concentrations, especially during the initial stage of infection and pathogenesis. Due to the high conductivity, and thermal and mechanical stability of CNT, it is considered a potential candidate for biosensor development, for instance, CNT-FET-based biosensors. However, the designing and simulating a high-performance, low-power, and miniaturized CNT-FET nanoelectronic device suitable for diagnostic applications, especially, point-of-care testing (POCT) is crucial for rapid and appropriate diagnosis of COVID-19 and other related viral diseases. Taking the leverage of the advancement of artificial intelligence, attempts have been made to boost the CNT-FET technology and the development of efficient CNT-FET-based biosensor models with accurate performance. This article explains the fundamental concept of the biosensor-based diagnosis of the COVID-19 disease, application of the artificial intelligence to increase the accuracy of the high-performance model, and the approach to standardize the design variables and performance parameters of the nanoelectronic circuit suitable for diagnosis. Moreover, the article highlights the current challenges and meaningful insights into their application in viral disease diagnosis beyond COVID-19 and the future perspective of the CNT-FET-based sensors in viral disease diagnosis.