What does an image prove?

This paper is a critical essay about image data that will attempt to uncover the characteristics that will allow images to be considered, among all ways or representation, as evidence. From two case studies where videosurveillange images are used, we have reflected on image data inflation, from its original role as a safety tool to its current role as a means of mass media entertainment. To explain the mutation of image data into evidence in a legal domain, we have used concepts of forensic aesthetics by Weizman, image spam by Steyerl and the genealogy of disciplinary images drawn from Tagg. By analyzing these case studies, we have found distinct features that we typically encounter in surveillance images that make them stand out from other contemporary images. Their vague appearance, their conceivable relation with the event, and the feigned absence of subjectivity ensure these types of images and their referent have a very solid connection, even to the point that they are considered the same thing. Image data, with its semiotic surplus value, causes complete confusion between representation and reality. Nevertheless, from our point of view, these images do not prove anything. They are merely a device to predict the past, and they serve essentially as a speculative area to project the wish to see what one wants to see. As experts in image production we have the duty to challenge this dynamic of representation and attempt at least to reveal the mechanisms of producing reality in the scopic regime of post-truth