Trustworthy AI for Human Behavior Understanding and Security

Recent advances in artificial intelligence have significantly enhanced the ability of intelligent systems to perceive, model, and predict human behaviors using multimodal sensing data collected from cameras, wearable devices, mobile platforms, and immersive technologies. These capabilities have enabled numerous applications, including human activity recognition, behavioral biometrics, user authentication, healthcare monitoring, and human-computer interaction. However, the increasing reliance on AI-driven behavioral analysis also raises critical concerns regarding the trustworthiness, security, privacy, and robustness of these systems. Emerging threats such as adversarial attacks, synthetic behavior generation, AI-generated motion trajectories, and data manipulation challenge the reliability of AI-based decision-making. Developing trustworthy AI frameworks that can accurately understand, authenticate, and safeguard human behavioral data has therefore become an important research priority across computer vision, cybersecurity, ubiquitous computing, and intelligent interactive systems.This Research Topic aims to bring together researchers from academia and industry to advance the development of trustworthy AI methods for human behavior understanding and security. We seek original contributions that improve the reliability, robustness, transparency, and security of AI systems designed to analyze, predict, and protect human behavioral data. The topic encourages interdisciplinary research spanning artificial intelligence, machine learning, computer vision, cybersecurity, behavioral computing, wearable sensing, extended reality (XR), and human-centered computing.We welcome contributions that develop novel algorithms, theoretical frameworks, benchmark datasets, evaluation protocols, and real-world applications for trustworthy behavioral intelligence. Particular emphasis is placed on bridging advances in human behavior understanding with security-oriented AI, including behavioral biometrics, continuous authentication, motion and trajectory analysis, predictive behavioral modeling, AI-generated behavior detection, adversarially robust learning, privacy-preserving machine learning, and explainable AI, etc. Through interdisciplinary collaboration, this Research Topic aims to advance trustworthy behavioral intelligence that supports secure, resilient, and human-centered intelligent systems across diverse application domains.This Research Topic welcomes original research articles, reviews, perspectives, benchmark studies, datasets, and application papers that advance trustworthy AI for understanding and securing human behavior. Relevant topics include, but are not limited to:- Human behavior understanding and behavioral analytics; - Human activity recognition and motion analysis; - Behavioral biometrics and continuous authentication; - Motion and trajectory prediction; - AI-generated behavior and synthetic motion detection; - Adversarial robustness and trustworthy AI; - Explainable and interpretable behavioral AI; - Privacy-preserving and federated learning for behavioral data; - Multimodal sensing and intelligent perception; - AI for wearable computing, smart environments, cyber-physical systems, and extended reality;- Benchmark datasets, evaluation methodologies, and real-world applications.Submissions presenting novel algorithms, theoretical advances, system implementations, interdisciplinary studies, and practical deployments are all encouraged. Topic editors: Mingjun Li, Sean Banerjee, Ashutosh Shivakumar, Xinchao Song Source: https://www.frontiersin.org/research-topics/84608/trustworthy-ai-for-human-behavior-understanding-and-security