Skip to content
Review Open access

Artificial intelligence, autism care, and health equity: a public health narrative review

Jul 2026 · Frontiers in Public Health · Vol 14 · 0 citations · 53 references
Medicine

TL;DR

It is argued that AI’s most important contribution to autism care is unlikely to lie in marginal improvements in classification accuracy alone, and its potential value lies in expanding access, supporting task-sharing, shortening diagnostic pathways, enabling population-oriented screening, and reaching under-recognised groups.

Abstract

Autism spectrum disorder (ASD) is a common, lifelong neurodevelopmental condition whose recorded prevalence, diagnostic delays, and uneven distribution of specialist services create a growing public health challenge. Conventional screening and diagnostic pathways depend heavily on scarce specialist expertise, contributing to long waiting times and unequal access across income settings, regions, sex, ethnicity, language, and social position. This narrative review synthesises current applications of artificial intelligence (AI) and machine learning in autism screening, diagnostic support, intervention, and longitudinal monitoring, and reframes the evidence through a public health and health equity lens. We argue that AI’s most important contribution to autism care is unlikely to lie in marginal improvements in classification accuracy alone. Rather, its potential value lies in expanding access, supporting task-sharing, shortening diagnostic pathways, enabling population-oriented screening, and reaching under-recognised groups such as girls and women, adults, ethnic and linguistic minorities, and populations in low-resource settings. At the same time, AI may create an equity paradox: technologies intended to reduce disparities may reproduce or amplify them if they are trained on non-representative data, deployed across a digital divide, or governed without adequate attention to privacy, accountability, and community trust. Whether AI narrows or widens autism-related health inequalities will depend on choices about data diversity, low-resource design, co-design with autistic communities, equity-sensitive evaluation, clinical integration, and proportionate regulation.

Read PDF

Similar papers

Open access Aug 2026

Autism care: reimagining the spectrum

Autism care policy is at a critical inflection point. Applied behavior analysis (ABA), long established as the “gold standard” through state insurance mandates in the US, has functioned as the default reimbursable intervention for autistic children. However, advances in genomics, neuroscience, developmental psychology, and scholarship on autistic lived experience have expanded understanding of autism as a heterogeneous neurotype characterized by meaningful differences in neural organization rather than a unitary disorder. Contemporary models emphasize neurodiversity, strengths-based perspectives, and the interaction between developmental processes and environmental contexts in shaping functional outcomes. Many autistic children also meet criteria for complex care needs, requiring coordinated, interdisciplinary services across health, educational, and community systems. This manuscript proposes reframing the “autism spectrum” from a hierarchy of symptom severity to a prevention-oriented “spectrum of care.” Adapting a public health taxonomy, interventions are organized into universal, selective, and indicated levels, targeting the prevention of avoidable disability, distress, and participation barriers. This model aligns autism services with whole-child, neurodiversity-affirming, and developmentally informed care, emphasizing relational health, autonomy, and life-course participation.

Steven Merahn · 1 citation
Review Jul 2026

From Data to Early Diagnosis: Artificial Intelligence as a Tool to Support Screening and Detection of Autism Spectrum Disorder in Childhood

Artificial intelligence shows promise as a supportive tool for early screening, but current evidence supports its use as a complement to, rather than replacement for, clinical assessment.

Andrea Catalina Mahecha Ballesteros, Juanita Valeria García Bello, Eleaine Scarlet González Zuñiga et al. · 0 citations
Open access Jul 2026

Healthcare Inequalities in Autism Services: Evidence from the United States, Africa, and High-Income Countries Using Machine Learning

This article presents a refined empirical framework for studying global disparities in autism spectrum disorder (ASD) diagnosis and service access across the United States, selected African countries, and high-income comparator countries. Unlike earlier drafts that treated the subject as a generic macroeconomic panel problem, the present manuscript is aligned with the accompanying data workbook, which identifies concrete sources for country-level ASD prevalence and burden estimates, U.S. Autism and Developmental Disabilities Monitoring (ADDM) Network surveillance measures, World Bank development indicators, WHO policy benchmarks, and optional clinical imaging data from ABIDE. The paper is written as a reproducible data-driven study rather than a claim of completed causal estimation: where the workbook provides templates rather than populated country-year values, the text distinguishes actual data sources from proposed estimands. The conceptual argument is that observed autism prevalence is not only a neurodevelopmental measure; it is also shaped by diagnostic infrastructure, clinical workforce capacity, school-based screening, insurance coverage, social awareness, stigma, income, and digital readiness. The proposed analytical design combines descriptive inequality indices, multilevel regression, Oaxaca–Blinder decomposition, random forest, gradient boosting, and SHAP-based explainability to identify the predictors most associated with cross-national differences in autism identification. The central contribution is a transparent, human-centered methodology for comparing autism diagnosis systems without overstating causal claims. The article emphasizes that lower reported prevalence in many African countries should not be interpreted as lower underlying need, because underdiagnosis, late identification, and limited surveillance capacity remain central measurement challenges. Policy implications focus on early screening, workforce development, culturally valid tools, telehealth, data infrastructure, and ethically governed AI systems for low-resource settings.

Irene Serwaa Agyapong, S. Atuahene · 0 citations
Review Open access Aug 2026

Challenges in Healthcare Access and Social Inclusion of Children with Autism Spectrum Disorder: Evidence from Kazakhstan and International Studies

Autism spectrum disorder (ASD) is a neurodevelopmental condition with a steadily increasing global prevalence. Despite advances in early diagnosis and intervention, children with ASD and their families continue to face significant challenges related to healthcare access, social inclusion, and stigma. These challenges are particularly pronounced in low- and middle-income countries but remain relevant worldwide. This narrative literature review synthesizes evidence on social, medical, and structural barriers affecting children with ASD, with a focus on healthcare access, comorbidity-related medical vulnerability, and public stigma in Kazakhstan and internationally. Databases including PubMed, Scopus, Web of Science, Google Scholar, eLIBRARY.ru, and KazNEB were searched for Russian- and English-language publications Access to ASD-specific services—such as early intervention, behavioral therapy, and speech and neuropsychological support—remains limited due to shortages of trained specialists, long waiting lists, and high out-of-pocket costs. Children with ASD frequently have comorbid chronic conditions, increasing their need for general healthcare, yet health systems are often poorly adapted to their sensory, communicative, and behavioral needs. Persistent stigma across social, educational, and medical settings further restricts access to care, while socioeconomic inequalities exacerbate disparities. Overall, children with ASD, particularly those with comorbidities, constitute a highly medically vulnerable population. Addressing these challenges requires integrated policy approaches, workforce development, expansion of publicly funded services, and targeted efforts to improve public awareness and reduce stigma.

Svetlana Mofa, Bauyrzhan Omarkulov, N. Delellis et al. · 0 citations
#explainable ai Review Aug 2026

AI and neuroimaging in autism spectrum disorder: advances in diagnosis, methodological challenges and future directions

A comprehensive review of recent advancements in ASD research, with particular emphasis on neuroimaging, artificial intelligence (AI), and machine learning (ML)-based diagnostic approaches, highlights the growing potential of AI-driven tools for supporting early ASD diagnosis and emphasizes the need for standardized protocols, external validation, explainable AI, and clinically translatable frameworks.

Kuljeet Singh, Khushi Mogha, S. Moctar · 0 citations
Open access Aug 2026

The Algorithmic Gaze: AI, Autism, and the Global Power Shift in Diagnosis

Artificial intelligence (AI) has rapidly emerged as a promising tool for improving autism spectrum disorder (ASD) screening, particularly in regions where access to trained specialists remains limited. Recent advances in machine learning have demonstrated encouraging diagnostic performance through the analysis of facial expressions, speech, eye gaze, and other behavioral markers, offering the potential to expand early diagnosis and reduce disparities in healthcare access. However, existing research has focused predominantly on algorithm development and diagnostic accuracy, while comparatively little attention has been paid to the broader ethical, cultural, and geopolitical implications of AI-assisted autism diagnosis. Drawing upon perspectives from philosophy of medicine and science and technology studies, this paper examines how AI redistributes power through three interconnected shifts: the transfer of diagnostic authority from clinicians to algorithms, the globalization of culturally specific definitions of "normal" behavior through Western-trained datasets, and the emergence of technological dependency on foreign-controlled AI infrastructure. Rather than rejecting AI-assisted diagnosis, this paper argues that equitable implementation requires culturally representative datasets, participatory model development, and greater local ownership of digital health technologies to ensure that AI promotes both diagnostic accessibility and global health equity.

Sophie Aoqing Qin · 0 citations