An Exploratory Document Analysis of Ethical Considerations in AI Policies and Guidelines: A Critical Review

Abstract

The rapid adoption of artificial intelligence (AI) across sectors such as healthcare, finance, law enforcement and education have increased the need for effective and ethically grounded governance frameworks. This study employs an exploratory qualitative document analysis to examine ethical principles embedded in 25 national and international AI governance documents published between 2020 and 2025. Guided by principle-based ethics, socio- technical systems theory and critical discourse analysis, this study investigates how ethical principles are articulated and operationalized across governance frameworks.

The findings reveal broad convergence around core ethical principles, particularly transparency, fairness, accountability, privacy and human oversight. However, significant differences exist in their interpretation and implementation. International frameworks predominantly rely on normative and aspirational language, whereas regulatory instruments increasingly embed ethical principles within enforceable compliance mechanisms. The analysis also identifies persistent challenges including ambiguous accountability structures, limited operational guidance, insufficient attention to marginalized populations and uneven consideration of emerging AI technologies. These findings highlight a gap between ethical commitments and their practical implementation across governance context.

The study contributes to responsible AI scholarship by highlighting the need for more accountable, inclusive and context sensitive governance approaches capable of translating ethical principles into practice