Akademik Veri Yönetim Sistemi
Dumlupınar Üniversitesi Sosyal Bilimler Dergisi, sa.89, ss.343-366, 2026 (TRDizin)
The speed with which artificial intelligence has proliferated in
organizational decision-making has intensified the accountability
crisis. In finance, healthcare, and human resources, AI systems
increasingly influence high-risk decision outcomes, with stakeholders
demanding transparency concerning how decisions are made and
accountability for failures. Yet prevailing understandings of
explainability are limited by siloed responsibility structures, weak
audit trails, and an under-specified “explainability burden”—the labor
associated with the production, maintaining, verifying, and assuring of
explanations for decisions made by AI. This conceptual article builds a
framework for distributing that burden among key stakeholders (AI
developers, data providers, process owners, auditors) using blockchain
as an immutable governance infrastructure. Building on research in
algorithmic accountability, institutional theory and distributed
governance, the article proposes a framework for quantifying explanation
expectations as a function of risk exposure and capacity and
implementing these allocations through smart contracts as smart
contracts on blockchain platforms. The model regards explainability as a
measurable burden that must be strategically allocated to ensure
transparency and legitimacy. I derive five testable propositions linking
blockchain-based auditability to accountability outcomes, stakeholder
trust, regulatory compliance and organizational learning, and provide
insights for organizations dealing with the EU AI Act, GDPR Article 22
and upcoming AI governance regimes.