How do technologies of computer vision, which promise to replace humans in the understanding of images, work in practice in the field of security, and what are their ethical and political implications?
Vision, understood as the capacity not only to see but also to make sense of what is seen, is increasingly being delegated to autonomous computer systems which influence how human operators determine suspicious behaviour. We currently lack an understanding of how these technologies impact governmental and private sector actors, their decision-making, and their accountability as well as the fundamental rights of those who are targeted. This project addresses these challenges through an innovative theoretical and methodological framework that investigates the theoretical, empirical and political implications of the development of computer vision in the field of security. In order to carry out this task, the project builds on and advances debates at the intersection of critical security studies, science and technology studies, and visual ethnographic practices.
Theoretically, the project attempts to overcome the debates in International Relations and critical security studies by offering a synthetic framework of analysis of socially embedded technical devices. Methodologically, the experimental use of creative coding and visual ethnography not only as a method of data collection and dissemination but also of data analysis within a multi-modal research design will advance, through visual practices, debates in International Relations about the status of non-propositional knowledge as well as alternatives modalities of presentation of research. Empirically it will bring a new understanding of the workings of computer vision technologies in the fields of social media content moderation of extremist content, “smart” Closed Circuit Television Vision (CCTV) cameras and “lie detectors” deployed at the border, participating thus to the ethical and political debates in the public sphere.