AI Trust: Can Explainable AI Enhance Warranted Trust?

Figure 1

Layouts of the mushroom game presented on the screen to the participants corresponding to each XAI model condition. In the mushroom game, participants could see the mushroom characteristics, a cartoon picture of the mushroom, and the AI recommendation. In the case of the two conditions with the presence of explanations, participants were also provided with explanations of the AI recommender system.

(a) Control layout

(b) LIME layout

(c) DICE layout