With GenAI now pervasive across media creation, archives need a shared language to guide their policies and to frame the implications for collections. Questions arise, for example, about when upscaling stops being a matter of improving quality and instead constitutes the creation of a new work, or how best to describe materials that are AI-generated or AI-altered in consistent and transparent terms.
The Taxonomies Working Group is addressing these challenges by gathering glossaries from across the field, analyzing existing frameworks, and identifying where gaps remain. Our goal is to create a comprehensive, field-wide metaglossary that enables archivists to clearly understand, evaluate, and communicate the range of machine-learning processes and their impact on archival media.