The Dignity Matrix of Mimetic AI Digital Twins

Mimetic AI digital twins reproduce recognizable features of a particular person. Their ethical evaluation therefore involves two human parties: the person represented and the person who encounters it. This paper proposes a two person dignity matrix evaluating each party independently. Drawing on Kant, dignity has a protective dimension, which prohibits reducing a person to a tool or commodity, and an enabling dimension, which recognizes a person as capable of choice and accountability. Crossing the dignity of the represented subject with the dignity of the user produces four categories: mutual respect, unauthorized continuation, authorized deception, and exploitative impersonation. The matrix offers a compact way to operationalize the ethics of dignity by identifying who controls the representation, who understands the encounter, and who is being used for another's purposes.