The Face Put to the Test by Artificial Intelligence – Between Presence and False Appearance

What becomes of the human face when confronted with artificial intelligence technologies that modify it, generate it, recognize it, and control it? This research brings together Emmanuel Levinas’s conception of the face, understood in the context of a new ethics, with an aesthetic approach to the face as a material that can be endlessly shaped by artificial intelligence technologies. Can an ethics of the face survive in a context that facilitates its manipulation? The aim is to examine how the ethical and aesthetic dimensions of the face, which tend to come into conflict in our uses of artificial intelligence, might be brought into relation without reducing one to the other.

First, the history of art shows us that the face has always been a technical surface that stimulates artistic creation and lends itself to transformation, without such metamorphoses necessarily posing a problem. Christ appears sometimes in the form of a young shepherd, sometimes as a fish, a phoenix, or even a dove. All these representations are considered legitimate. None claims to reproduce Christ’s exact appearance. Each is, so to speak, faithful to him. By contrast, advertising makes instrumental use of the face with the aim of seducing the consumer. The face serves only to enhance the product, without signifying in and of itself. At the same time, theories such as physiognomy encourage us to regard the face as a grid for reading that provides access to another person’s interiority. They assume that it is possible to control anyone on the basis of their face. This reduction of the person to a metric is intensified by practices involving artificial intelligence, such as deepfakes and AI-generated images. These practices have real consequences for individuals’ lives: identity theft, fraud, and political propaganda are some examples. Emmanuel Levinas proposes moving beyond this aesthetic relation to the face in order to preserve the life it embodies.

Second, in his work Totality and Infinity (1961), Emmanuel Levinas draws on the concept of the face with the intention of placing alterity at the heart of his ethics. The philosopher describes the face of the Other as beyond grasp: although the materiality of the face is finite, presence is infinite. The interiority of the Other remains inaccessible, making it impossible to know or control the Other completely. Yet the face speaks and expresses itself. It articulates a discourse, even in silence, and enables individuals to enter into relation without merging or becoming indistinguishable from one another. The face says, “Thou shalt not kill,” because murder signifies both the ultimate desire to exert total control over the Other and the failure to which this undertaking necessarily leads, since once the Other is dead, no relation is possible. This prohibition against murder is rooted in the very vulnerability of the face exposed to the violence of the Other. Because I can reach and destroy the Other’s face, because it is vulnerable, the face imposes upon me an unlimited and unconditional responsibility that precedes my freedom. It is an asymmetrical obligation, insofar as it rests neither on reciprocity nor on rational calculation. It is a call from the Other, a demand that can never be discharged, because it constitutes the fundamental structure of subjectivity. This demand always exceeds what I actually do. But what ethical future does the face have in the age of artificial intelligence?

Third, artificial intelligence technologies serve both economic and political objectives aimed at the alienation of the face, through facial recognition, while at the same time offering a space in which the face can be reinvented. Can an artificial face exist? Science fiction teaches us that the face, in the Levinasian sense, is not the exclusive prerogative of the human. If artificial intelligence can simulate a presence and retrieve personal and sensitive data for obscure purposes—as illustrated by the criticism directed at the connected necklace developed by friend.com—it might also embody a new form of alterity in a more or less distant future. In the immediate present, our uses of artificial intelligence reactivate a dangerous mechanism of distancing and disengagement from responsibility, based on the delegation of tasks, which the historian Johann Chapoutot calls “intermediation.” Thus, the mediation of interactions through digital interfaces creates the impression of proximity without requiring commitment. Reduced to an image in circulation, the face loses some of its ethical power. Can we then conceive of a non-neutralizing aesthetics? When integrated into artistic, educational, and critical practices, AI could help reveal the face, and through it alterity, without confining it.

In the context of the emergence of artificial intelligence, the face—already omnipresent in the visual landscape—is both a privileged target of abuse and potentially a site of resistance, from which to subvert the dominant aesthetics and reaffirm the presence of alterity. By placing the face at the foundation of his ethics, Emmanuel Levinas reminds us of the individual and collective responsibility that falls upon us at all times and in all places. New technologies do not necessarily herald the disappearance of the face, but they compel us to rethink the conditions of possibility of an ethical encounter.