This chapter explores human-centered artificial intelligence (AI) as a paradigm that integrates human needs, capabilities, and ethical considerations into AI-powered cybersecurity systems, aligning with the science of security’s emphasis on systematic, interdisciplinary solutions. It examines AI’s dual role in enhancing both defensive and offensive cybersecurity capabilities, and highlights the importance of human users in both. The chapter addresses critical challenges such as data quality, inherent biases, privacy issues, and the demand for explainability and transparency in AI-powered cyber defense. The authors propose using human-centered approaches to address these challenges, including the development of explainable and trustworthy AI-powered cyber defense systems, human-centric design, user training and education, effective human-AI teaming, bias mitigation, and supportive regulatory frameworks. Case studies, including cybersecurity for autonomous vehicles and defenses against AI-enhanced phishing attacks, further illustrate the practical applications and dual-use nature of AI in cybersecurity. The chapter concludes by envisioning a future where human-AI collaboration, grounded in scientific rigor, ethical standards, and regulatory frameworks, with interdisciplinary efforts, ensures robust and trustworthy cybersecurity solutions.