Social Construction of XAI: Do We Need One Definition to Rule Them All?

Different stakeholders need different explanations: The paper argues that researchers, developers, regulators, and end-users all have legitimately different interpretations of what "explainability" means based on their needs. Rather than forcing one definition, the field should recognize this "interpretative flexibility" as a feature, not a bug.

Premature standardization could harm innovation: Rushing to establish a single, universal definition of XAI risks stifling creative approaches and locking in assumptions that may later prove limiting. The paper suggests the field is still too early in its development to benefit from rigid standardization.

Pluralism can coexist with progress: The authors contend that XAI can advance meaningfully without waiting for definitional consensus. Different communities can work with their own interpretations of explainability while still making practical improvements to AI transparency and trustworthiness.

Social context shapes technology meaning: Drawing on Social Construction of Technology theory, the paper demonstrates that what "explainability" means is not purely technical—it's socially constructed through interactions between different groups with different interests and priorities.

One-size-fits-all approaches are unrealistic: The paper implicitly warns against regulatory or industry efforts to impose a single definition of XAI across all contexts, suggesting such approaches would fail to account for the legitimate diversity of needs across different applications and stakeholders.