From Extended Self to Posthuman Other: Empathy and Moralityin Encounters with AI Characters

This paper explores the entanglements between morality, empathy, and artificial intelligence, asking how contemporary videogames mobilize different forms of empathy in their representation of AI characters.Introducing a taxonomy of three key configurations of representing AI characters, we show how videogames respond and adapt to cultural imaginaries of AI.While most research to date focuses on players' bond with the avatar (e.g., Wilde 2024), the focus of this paper lies with non-playable characters (NPCs).Here, we see a strong connection between empathy-an affective and/or cognitive perspective-taking of another's position-and morality, understood in this paper as moral intuitions, an intrinsic feeling of 'right' and 'wrong' based on one's emotional engagement (Joeckel et al. 2012).The connection between morality and empathy is a bone of contention across disciplines.Proponents of empathy see it as a distinctly pro-social skill that fosters a more positive and reflective engagement with diverse life realities (Smith 2011; Song 2015).Conversely, critics of empathy highlight how our imaginative engagement with morally flawed or even downright evil characters can invite us to identify with their motivations and downplay or justify their wrongdoing (Vaage 2023).In game studies, important criticism has been raised against the notion of games and VR as "empathy machines," pointing to their tendency to appropriate and commodify Othered experiences (Nakamura 2020; Ruberg 2020; see also Foxman et al. 2021).The addition of AI accentuates these frictions in two interrelated debates, which, rather interestingly, link questions around the moral status of AI to its empathic relationship with humans.The first debate is concerned with the potential status of AI as moral agents, which many scholars deny precisely because AI systems lack