A virtual reality and AI-based system can detect autism in young children with over 85% accuracy, surpassing traditional methods.The system observes children's motor movements and gaze patterns in immersive virtual environments for more naturalistic responses.It uses deep learning to identify behavioral biomarkers linked to ASD, enabling efficient and affordable diagnosis.The system's approach could enhance early autism detection accessibility and pave the way for studying other motor symptoms in ASD.Researchers at the Human-Tech Institute-Universitat Politècnica de València developed this innovative detection system.The system projects a child's image in simulated environments to analyze movements and behaviors for accurate diagnosis.A newly developed AI model in the system outperforms traditional AI techniques, improving precision in ASD detection.The Human-Tech Institute team's eight years of collaboration has focused on improving early ASD detection through virtual reality.The system shows promise in the analysis of motor activity as a biomarker for autism detection.Future adaptations of the AI model could extend to analyzing ASD patients' movements in various tasks.