Our everyday life is increasingly permeated with digital objects that carry out smart and complex functions. The epilogue is to be identified the creation of a field, at once conceptual and material, of things denominated smart objects (henceforth SOs). These objects are in a sense the “population” of what is defined as a new development phase of the internet/web “system”: the Internet of Things (IoT). This technological evolution is so pervasive that it is referred to as smartification, a process which like other cultural phenomena – as in the case of gentrification – strongly characterizes social action. It is evident that the phenomenon is by no means easy to define: which objects are really smart and which are not? But above all: what do we mean in semio-linguistic, and not psychological nor merely phenomenological terms, when we attribute the qualifier smart to an artifact? What is clear is that a new, or at least different (and controversial) relationship is developing between objects and subjects, or rather between human beings and objects inhabiting the spaces of social action: that is, a new system of objects, to cite a “classic” by Baudrillard (1968), or a new “society of objects” (see Landowski, Marrone 2002). In this presentation we will focus on a type of device designed to interact with its users in the domestic sphere, assisting them in a variety of tasks – such as for example Amazon’s recently developed Echo, capable of connecting to Alexa, an intelligent personal assistant also developed by Amazon’s programmers and based on machine learning of which millions have been sold in the USA, or the more recent Google Assistant. Our semiotic-oriented – or more precisely socio-semiotic and virtually ethno-semiotic – analysis will deal with these issues by concentrating on the problem of identity, which is anthropologically, but also and above all philosophically, sensitive in this context (but that has a long-standing tradition of critical discourse in semiotics too, from Saussure onward). Our perspective, indeed, aims to evaluate the extent to which human enunciating subjects, in interfacing discursively with SOs, turn them into agentive simulacra – thus into inter-actants of their discourse, non-human (nor animal) im-personal actors: a sort of “disincarnate” partners in a performance in which they act as real co-enunciators, not mere projections or prostheses; and all this without necessarily projecting onto SOs the image of conscious cognition, or attributing a “personal” (if not even intentional) identity to them. In our analysis, we will take into consideration only what may be defined as the agentivity of SOs within an expressly non human-centric perspective (that is, not built around a biological subject as the only real actant, as in the nonetheless innovative theories of embodied and embedded cognition). Dealing with the concrete actions of the human and non-human actors concerned, it is possible to observe and define the plane of relations that produce signification and which is created within the multiplicity of enunciative instances. Our main hypothesis is that SOs can lead to a comprehensive revision in the possible versions of the concept(s) of identity which can be formulated within disciplines concerned with the study of semiotic systems.