Exploring Authentic Anonymity in Documentary Film through AI
Summary
An artistic research project exploring how AI can anonymise real people in documentary footage while preserving authenticity and credibility. Traditional anonymisation techniques — heavy blurring, pixelation, strong voice distortion — introduce stigma often associated with victimhood or criminality, associations that are misleading when used to anonymise people who require anonymity for other reasons. At the same time, these techniques remove what is central to the story: our ability to read body language, facial expressions and tone of voice in order to understand human emotions and relationships. Drawing on extensive video recordings from an ongoing artistic research project on roleplay and embodied methods in documentary education at the TV School in Lillehammer, and with MishMash partner Reimagine as key collaborator, the project will test existing AI tools and develop new models that transform participants’ identifiable features while preserving their expressions, emotions, interactions and environment, in line with ethical and legal requirements. Outcomes include an overview of viable approaches, a preliminary anonymisation workflow, and a plan for a larger project towards a publishable video essay.
Work Package
Primary WP: WP2
Duration
2026
People
- Project leader: Mette Nyseter