Industry and society are experiencing the transformational impact of the autonomous systems revolution, empowered by automation capabilities offered by Artificial Intelligence (AI). Cyber-physical Systems of Systems (CPSoS) define a multi-faceted and dynamic environment where autonomy is fundamental to govern the complexity of interactions between the virtual and physical worlds with minimal human intervention. However, even when the most advanced degree of autonomy is exercised, the human is a variable which cannot be left out of the CPSoS equation, particularly in safety critical scenarios like autonomous transportation. TEACHING puts forward a vision of humans at the centre of autonomous CPSoS, by embracing the concept of Humanistic Intelligence, where the cybernetic and biological entities cooperate in a mutual empowerment towards a shared goal and where human feedback becomes a crucial driver for CPSoS adaptivity. TEACHING addresses the challenge by integrating AI with fundamental concepts of security and dependability stemming from the AI-human-CPSoS interactions, and by considering their impact on the underlying computing system. TEACHING develops a human-aware CPSoS for autonomous safety-critical applications, based on a distributed, energy-efficient and dependable AI, leveraging edge computing platforms integrating specialized computing fabric for AI and in-silico support for intelligent cybersecurity. The goal is to design a computing software and system supporting the development and deployment of adaptive and dependable CPSoS applications, allowing to exploit a sustainable human feedback to drive, optimize and personalize the provisioning of the offered services. TEACHING outcomes will fundamentally impact the development of autonomous safety-critical systems, providing means to improve their safety, dependability and overall acceptability. This impact will be demonstrated by TEACHING in two pilots concerning autonomous driving and aviation.