Does Explainable AI Have Moral Value?

XAI is fundamentally a moral duty, not just a consumer right. The paper argues that explainability should be understood as part of our ethical obligations to people affected by AI systems, rather than simply a feature users can demand. This reframing emphasizes that organizations have a responsibility to explain AI decisions because it helps maintain fairness and human agency in decision-making processes.

Explanations help preserve human dignity and equality in AI systems. According to the paper, XAI sustains "constitutive symmetry and agency" in AI-led decisions—meaning explanations help ensure that people aren't treated as passive subjects of algorithmic judgment but remain active participants who can understand and potentially contest decisions affecting them.

Current XAI efforts fall short of their ethical potential. The paper identifies significant gaps between the ideal of reciprocal relationships and what's actually being implemented: machine learning focuses on technical methods while ignoring real-world adoption challenges, human-computer interaction oversimplifies how organizations actually work, and policies demand accountability without providing technical guidance.

Different AI communities are working in silos without a unified ethical vision. The paper reveals that machine learning experts, interface designers, and policymakers each approach XAI differently and incompletely, suggesting that responsible AI requires these groups to synthesize their perspectives rather than operate independently.

Making XAI truly ethical requires addressing practical implementation barriers. Beyond developing better explanation techniques, the paper emphasizes that responsible AI demands solving organizational, technical, and policy challenges to ensure explanations are actually meaningful and accessible to affected people.