Deep fakes as a tool of political advertising. Can regulatory framework benefit from the “Ship of Theseus” paradox?

Deep fakes have been consistently perceived as a threat to social and political processes, particularly elections. A rarely discussed phenomenon, though gaining importance considering the increasing number of real-life use cases, is using deep fakes for political advertising purposes. By manipulating visualizations, artificial intelligence (AI) allows politicians to eliminate or minimize the deficits of their performance and highlight or create features perceived as attractive by the audience. Such manipulations pose a significant epistemic problem, and can severely distort perception of real politicians and their characteristics. These disruptions might correspond to the classic “Ship of Theseus” paradox that addresses the problem of changing identity. The aim of this study is to analyze how this philosophical conundrum could be used for crafting better transparency obligations (disclaimers) for using deep fakes for political advertising purposes. The aim of such disclaimers is to sensitize and protect public opinion against manipulation. However, at the moment the basic labels concentrate on technological dimension of manipulation, overlooking cognitive consequences for human audiences. This paper proposes extensive disclaimers to strengthen their informative and sensitizing potential.