Culturally Adaptive AI in Autism Screening: A New Paradigm for Equitable Early Detection

Early detection of autism spectrum disorder (ASD) is crucial for timely intervention and improved outcomes. While artificial intelligence (AI)-based digital screening tools have enhanced the objectivity and scalability of autism diagnosis, they often fall short in addressing the diverse cultural and linguistic contexts of underrepresented populations. This review highlights the emerging need for community-driven, culturally adaptive AI screening systems that prioritize inclusivity and equity. By involving local communities in the design, data collection, and validation processes, these systems can better capture culturally specific behavioral markers and overcome barriers related to language, literacy, and access. We discuss current AI screening tools, their limitations in diverse settings, and propose a research framework that integrates participatory design, localized AI training, and ethical considerations. Implementing culturally responsive AI tools within community health infrastructures has the potential to reduce diagnostic disparities and empower families worldwide. This paper aims to inspire future research and collaboration toward developing autism screening technologies that are not only accurate but also culturally meaningful and accessible to all.