Artificial Intelligence (AI) Misuse: Perpetrators’ Moral Justifications

Artificial intelligence (AI) misuse, like creating deepfakes, can have severe consequences for victims, such as damaging their reputations. Extending prior AI misuse research focusing on the victims’ perspective, we focus on the perpetrators’ perspective to explain their AI misuse. We draw on moral disengagement theory (MDT), suggesting that such unethical behavior is subject to perpetrators downplaying their behavior and directing the responsibility to others. Considering the unique characteristics of AI, such as human likeness, we contextualize MDT by considering that perpetrators may direct the responsibility for AI misuse to the AI itself. We will develop valid measures for this novel moral disengagement dimension in a planned first study and integrate it with the established MDT to explain AI misuse in a planned second study. We will contribute by explaining why perpetrators misuse AI, contextualizing moral disengagement theory to AI misuse, and showing if and how the moral disengagement dimensions interact.