In the age of AI, the photographic image raises not only questions about the medium itself, but also broader cultural concerns. Can it, in this form, still bear witness to photography’s temporal dimension – to the presence/absence of lived reality, as defined by Roland Barthes – as well as to authorship or authenticity? How should we approach the further saturation of an already overpopulated digital space, and the significance of art and visual culture more broadly? Tena Lovrenčić explores these questions.
Isidore Hibou, “Impression, soleil levant” (detail), 2024, as part of the exhibition “AImagine – Photography and generative images” at the Hangar Gallery in Bruxelles. Photo by: Sandra Križić Roban
Art, as a socially conditioned form of creative practice, has never existed in isolation, but has always developed alongside technology. Their relationship has often prompted a redefinition of established modes of artistic production, presentation, and perception. Photography is a prime example of a technological innovation that transformed the traditional paradigms of the art world in the mid-nineteenth century. When photography first emerged as a new medium, it sparked considerable controversy, and many critics, writers, and artists proclaimed the imminent demise of the (visual) arts. Today, however, it is widely acknowledged in art-historical discourse that the opposite occurred: photography freed painting from the demands of mimetic representation and the “replication” of visual reality, paving the way for experimental approaches and exploration of new forms of artistic expression.
Because of its ontological specificity and multidimensional character, photography continues to pose theoretical challenges even after establishing its status as an art form, and often occupies a central place in critical discourse. Each new wave of technological development – from the daguerreotype and the transition from analogue to digital photography, to the broader virtualisation of the medium – has generated new debates in photographic practice and theory. Many rightly argue that the photographic image exists in a state of constant crisis or, at the very least, continuous transformation. Changes in the technical and technological aspects of the photographic process not only open up new expressive possibilities for the medium, but also broaden the interpretative field of the photographic image and its semantic revalorization.
Generative artificial intelligence (AI) is one of the key technologies underpinning the Fourth Industrial Revolution. In recent years, we have been confronted with a proliferation of sensationalist headlines and debates surrounding the latest “horseman of the apocalypse”, as AI is rapidly integrated into every sphere of public and private life – including visual culture and the arts. AI-based systems such as Midjourney, Stable Diffusion, and DALL-E 2 have demonstrated an ability to generate content that, according to certain criteria, can be perceived as creative. Generative AI systems operate on the basis of textual prompts, drawing on vast datasets of images to generate visual content.
The automation of the creative process and the rapid advancement of generative AI systems raise a series of questions within the art world. Many draw parallels with the reactions that accompanied the invention and emergence of photography, although the two are fundamentally different phenomena with distinct trajectories. Debates centre on questions of authorship, ownership regulation, the democratisation of the medium, plagiarism and appropriation, hyperproduction, the transformation of the art market, and the broader reshaping of artistic processes and curatorial practices. Just as the advent of photography once prompted proclamations of the death of painting, AI is now giving rise to narratives about the possible end of photography – and even of art as such.
One of the first viral examples of the use of artificial intelligence in photographic practice was Boris Eldagsen’s THE ELECTRICIAN, from the PSEUDOMNESIA series, created in 2023. The term pseudomnesia refers to a false or distorted memory of events that never actually occurred, which perfectly reflects the synthetic and artificial nature of AI-generated visual content. Drawing on the distinctive visual language of 1940s portrait photography, the work presents a fabricated transgenerational memory that, as the title suggests, is essentially a hallucinated recollection of a non-existent past. According to Eldagsen, it is a past that no one photographed and that was never photographed – a past that never existed in the first place.
This brings us to the concept of temporality discussed by Roland Barthes in his seminal work Camera Lucida: Reflections on Photography. Barthes regards photography’s temporal dimension as one of its fundamental qualities, linking it directly to the phenomenon of memory. For him, the photographic image is a certificate of presence, proof that something really happened, a testimony to a subject that existed in a particular time and place. At the same time, it is a confirmation of absence: the moment that existed when the photograph was taken is now irretrievably lost. The uniqueness of the photographic medium lies in this very paradox. Photography functions as a kind of memento mori; even as it freezes time, it points to its inevitable passing.
The inherent and inextricable connection between photography and the real world is a marker of its authenticity, in contrast to AI-generated visual content, which has no referential relationship to reality. Promptography – a neologism used to describe content generated by artificial intelligence – is devoid of temporal and spatial context; it presents a fictional simulacrum within an artificially constructed space and time. Although photography, in all its forms, is only an interpretation of reality, its documentary potential continues to be central to photojournalistic practice, where it functions as a factual visual record. Because of its capacity for manipulation, as well as its tendency to distort reality and familiar visual markers, artificial intelligence poses a threat to the very concept of truth, making it imperative to address the ethical questions it raises and to regulate its use.
The question of authorship, which recurs throughout the history of art, concerns the issue of who is the actual author of a work – in other words, to whom its creative or intellectual ownership should be attributed. In the context of artificial intelligence, this question becomes particularly complex. Who is the author? The AI itself? The programmer who developed the model? The person who entered the prompt?
A photographic image does not simply materialize out of nowhere; it is the result of a whole series of actions and authorial decisions. These range from technical and technological choices – such as the type of camera, format, exposure, and perspective – to the deliberate act of framing a particular scene. The photographer’s subjectivity and individual “signature” emerge precisely through the act of framing – the conscious selection of fragments from visual reality – which forms the basis of the creative process.
AI-generated content adopts the visual elements and conventions of photographic language, but the process through which it is produced is not equivalent to photographic practice. Generative AI can imitate the photographic image with striking efficiency and hyperrealism, but it annuls the author’s subjective experience and control over both the process and the final result. The traditional role of the author is thus radically transformed, as they are no longer the creator of content, but rather its modifier and selector. The significance of art as an intrinsically human activity, and of processuality as an integral part of creative work, is being devalued by mechanisation as never before. In an age of automated interpretations and generated realities, it is becoming increasingly difficult to discern what is original and authentic.
Given that artificial intelligence uses datasets composed of original works by real artists, the images it generates can be seen as appropriated derivatives of those works, raising the issue of fragmentary plagiarism. Beyond the evident ethical and legal concerns, it is reasonable to anticipate increasing homogenisation and a decline in quality resulting from the limitless production of copies. Cyberspace, already saturated with content, risks becoming an even larger repository of digital waste, while the intuitive anxiety and collective fear surrounding AI’s monopolisation of visual culture are understandable and entirely justified.
To ensure that artificial intelligence does not become – or remain – merely a destructive catalyst for change, it is necessary to establish a balance between the use of new technology as a tool and the preservation of a unique authorial vision.
As part of the PhotoBrussels Festival, earlier this year the renowned Hangar Gallery hosted the first group exhibition of works created using artificial intelligence titled AImagine – Photography and Generative Images. The curatorial concept sought to present the possibilities of AI-generated imagery, explore the intersection between generative images and photography, and establish a dialogue between tradition and innovation with the aim of reinterpreting and deconstructing familiar visual codes. The potential of artificial intelligence in artistic practice is perhaps most evident in conceptually driven works that move beyond the dichotomy between fiction and reality, thereby expanding the artist’s perspective. Research-based, analytical, and reflective works created with the aid of artificial intelligence can be particularly compelling when they critically examine the technology itself.
Responses to the integration of AI-generated content into visual culture – whether virtual or physical – range from dystopian to techno-enthusiastic. The scope and repercussions of generative AI remain largely speculative and uncertain, but they undoubtedly open up a hybrid space that will attract artists interested in experimental methods, multimedia practices, critical approaches, and research. In an age of automation and posthumanism, as the boundary between the fictional and the real gradually dissolves, it is important to remember that there is always a counterpoint – a space of resistance, discourse, complexity, and new possibilities.
Tena Lovrenčić
(translated by Tihana Bertek)
