Emotion-sensitive Explanation Model

Emotions significantly influence how people understand AI explanations. The paper challenges the traditional assumption that XAI should focus solely on rational understanding, demonstrating that "prior and task-generated emotions can negatively impact the understanding of explanation," meaning users' emotional states directly affect whether they comprehend AI systems' decisions.

A three-stage model connects emotions to decision-making outcomes. The proposed framework identifies emotional arousal, understanding, and agreement as interconnected stages, suggesting that effective explanations must account for users' emotional states at each step rather than treating emotions as obstacles to overcome.

One-size-fits-all explanations are insufficient for responsible AI. The paper advocates for "XAI systems that dynamically adapt explanation strategies to users' emotional states," implying that responsible AI deployment requires personalized approaches that recognize different users may need different types of explanations depending on their emotional context.

Emotion-aware design supports better user decision-making. By acknowledging emotional factors, the model aims to support "more effective and user-centered decision-making," suggesting that AI systems designed with emotional sensitivity may lead to more trustworthy outcomes and better alignment between AI recommendations and human values.