Implicated Images, (2026)
Implicated Image #01–#13, 2026. Digital C-prints, 121.9 × 96.5 cm | 48 × 38 inches each
In Implicated Images I photographed unsecured, publicly accessible security camera feeds. Those who appear in the images are presumably aware of the presence of security cameras but did not know I was watching. Similarly, they do not know you are looking.
Security cameras deal in what I call predictive images. They are composed not in the present, around a subject or a circumstance, but around potential -something will happen. These unusual images carry no weight until they witness something of significant, circumstantial evidence.
As a result, literacy of these images comes with the expectation of occurrence, usually unfortunate. Through observation, the subject is damned to the consequence of event and often the viewer implicated in their attention with no potential to intervene.
Further, photographic associations to Henri Cartier-Bresson’s decisive moment inform that something is about to happen on screen. Otherwise, why would these photographs be presented? Like the tension inside a bending branch, it is bound to snap eventually.
However, these photographs are non-events. Their context in the subject of our gaze
leans towards the potential futures of something around that corner, who this person is, and what is about to happen. Yet there is no resolution. The photographs remain unfinished.
Security photographs are operational images intended for machine vision. These cameras do not see from a human perspective -which in turn made these complicated photographs to view and to make.
I had to remind myself of this when photographing -some may seem unsettling or potentially violent when understood through a human perspective. They are not, insofar as the machine is concerned. The machine makes no distinction between what is worth recording and what is not. To truthfully attempt to convey this in translation to a photographic image, I could not censor through my own associations or discomforts.
Although the photographs depict events unfolding somewhere in a real, tangible
space, their primary location is not geographical but computational. Before becoming photographs, they exist solely as compressed streams of data moving through network infrastructures. These live feeds did not offer a way to determine a precise location of occurrence. Watching was not a choice that I made over intervening -it was the only action the image permitted.
Some images required only minutes before a tension emerged. Others required hours or repeated observation across multiple days waiting for subjects. I locate a composition and remain within it, allowing chance to determine what may eventually enter the frame, akin to traditional photography techniques like those of Cartier-Bresson who often would determine his composition geometrically around the potential of a scene and wait for the decisive moment to enter the frame.
However, in acknowledging the difference between photographing a physical location in a more traditional street photography methodology versus a digital image landscape, Implicated Images looks to the work of artists Jon Rafman and Michael Wolf, both of whom have explored imagery produced through machine vision and networked image systems. Rafman’s The Nine Eyes of Google Street View (2008–ongoing) navigates Google’s Street View by recontextualizing screenshots, while Michael Wolf’s A Series of Unfortunate Events (2010) rephotographs scenes discovered within Google Street View using a traditional camera.
The distinction lies in temporality. Rafman and Wolf use found images from what may be considered archived images that are suspended in a digital record. Their photographs emerge through acts of searching and recontextualizing images that already exist within Google’s mapped database. Google Street View presents a completed and mapped past of a landscape. Surveillance imagery from live feeds presents a predictive future. Every photograph in Implicated Images originates from a live surveillance feed. The events depicted unfolded in real time while I watched.
Farocki called these operational images -images made for machines, not for viewers. Operational images derive from military imagery and depart from the responsibility or expectation of an image to represent the world rather than influence perception. Farocki observes they are produced not for human viewing but for technical systems that guide machines, monitor environments, or automate decisions. These are images which are functional before they are visual. Jussi Parikka expands on Farocki’s idea of the operational image, emphasizing that operational images do not stop at mere observation. They actively interpret the environment, often through images processed by algorithms instead of human perspective. This understanding suggests that my intervention in presenting Implicated Images is not necessarily the creation of the image itself but its translation into a photographic object.
My recontextualization through interventions with scale and material is an effort to consider our relation to and understanding of these non-native images. Security images are pseudo. They resemble something familiar and legible. However, under a closer look, nothing is truly discernible. There is no context to be taken, no sure identities, and nothing besides assumptions and associations from which we have projected.
These photographs resist interpretation.