Synthetic Photography is a concept developed in Kevin Abosch’s practice and is clearly distinguished from simple AI image generation based on prompts. Drawing on his background as a biologist, as well as earlier experience with laboratory data analysis and coding, Abosch trains diffusion models using tens of thousands of photographs from his own archive and additional self-collected datasets. In this framework, model-building is not treated as an automatic image-output process, but as a dialogic act of sculpting a machine’s internal visual understanding of the world.
The concept is also tied to questions of photographic truth. In works such as #CIVICS and Ethical Work, Abosch presents synthetic photographs that initially resemble documentary or press images, yet contain visible AI-generated noise when examined closely. These works stage protests, civil disobedience, and police pressure in cities around the world, and they directly address the crisis of post-truth in contemporary society. By doing so, Synthetic Photography asks where the boundary lies between reality and fabrication, and how images come to be mistaken for truth.
Abosch’s approach has also produced friction with traditional art institutions and critics, who have accused such works of undermining photography’s truth value or merely imitating reality. However, the confusion and anger generated by the question of whether an image is real or fake are themselves treated as part of the work’s completion.