Modern digital systems are increasingly pervasive and autonomous, also due to the widespread use of AI technologies. They concur to create socio-technical systems, composed of distributed entities interacting with and within the environment, under constant perturbations. This resembles of Collective Adaptive Systems. However, from the human-perspective, AI-enabled autonomous and adaptive systems raise more than just privacy concerns; they also affect other ethical values and human dignity. In this paper, we introduce the concept of ethical-aware Collective Adaptive Systems and we outline challenges to design Collective Adaptive Systems respectful of human ethical values.