The decisions, forecasts and operations produced by intelligent systems reveal new possibilities for action in various areas of society and radically transform our lifestyles. This epochal change has triggered, in parallel with a remodeling of our habits, an important process of re-semantisation of some notions traditionally ascribed to human beings, such as those regarding intentionality and responsibility. This paper addresses the process of redefining these notions with the aim to shed light on composite forms of intentionality and responsibility in the increasingly close relationship between humans and AI. To this purpose, we propose to apply in the specific field of AI the idea of composite intentionality developed by P.P. Verbeek and extend it to the notion of responsibility. Despite some terminological problems related to Verbeek’s proposal, his post-phenomenological approach has the merit of taking the human-technology association as its main focus, helping us to better understand the moral status of artificial intelligence.