To be trustworthy and to trust: The new frontier of intelligent systems

The new frontier of intelligent systems is their ability to enter into advanced collaboration with humans and other more or less intelligent artificial systems. This collaboration is based on the autonomy of artificial systems and their ability to solve problems in multiple domains as an evolving process that represents an extraordinarily useful perspective as long as we can trust their finalities and uses, their implementation, and their behavior. For several years, the domain of trustworthy AI and trustworthy Intelligent Systems has been extensively studied, in particular to make these systems safe and prevent them from damaging human activities and, at times, human beings themselves. However, there is another dimension that, if properly developed, allows intelligent systems to play extraordinarily useful roles and is given by their ability to trust other intelligent autonomous artificial systems or humans themselves. This dimension is fundamental for the development of truly cooperative and effective relationships and activities. In this work, we introduce a systematic review of the state-of-the-art by providing a broad overview of the existing work on trust in intelligent systems. In particular, we investigated three macro domains where trust had and currently has a crucial impact: IoT, MAS, HMI.