<p>Autism spectrum disorder (ASD) is a complicated neurodevelopmental disorder that affects a person’s ability to communicate socially. Early detection of ASD is necessary for effective intervention, but conventional diagnostic methods that depend on behavioral evaluations are difficult and unreliable. The goal of this systematic review is to identify the use of the eye tracking method and machine learning (ML) model for early ASD detection. This review study also identifies significant advancements, difficulties, and future directions for this field of study. To identify relevant studies employing the eye-tracking method, this review was conducted in accordance with the PRISMA guidelines. A comprehensive search of PubMed, Pubmed Central, IEEE Xplore, and Google Scholar from 2020 to 2024 was conducted due to their broad peer-reviewed coverage across disciplines relevant to eye tracking and ASD. Out of 1006 articles discovered throughout the search process, 35 peer-reviewed papers that satisfied the inclusion and exclusion criteria were chosen. This review shows how ML models with high diagnostic accuracy can be used to improve eye tracking as a safe, objective method of detecting unusual gaze patterns in ASD. This study also shows several challenges in the reviewed studies, including small sample sizes, limited demographic diversity, and device variability. This study suggests more diverse datasets, standardized methods, and affordable technology to increase the clinical adoption of the technology for future research. This study also contributes to the field by offering a comprehensive evaluation of existing approaches, guiding future researchers in the direction of more accurate and effective early autism diagnosis methods.</p>