Toward Effective AI Governance: A Review of Principles

Transparency and accountability are the dominant principles in AI governance frameworks, appearing most frequently across reviewed literature, though the paper notes these principles often lack clear implementation mechanisms for organizations to follow in practice.

The EU AI Act and NIST Risk Management Framework (RMF) are the most influential governance frameworks currently shaping the field, suggesting that regulatory and technical standards are leading the conversation around responsible AI development.

There is a significant gap between principles and practice: while many governance frameworks emphasize ethical values, the review found that "few reviews detail actionable governance mechanisms or stakeholder strategies," meaning organizations often struggle to translate high-level principles into concrete policies and procedures.

Current AI governance literature lacks empirical validation and inclusivity: the paper identifies gaps in how well these frameworks are actually tested in real-world settings and notes concerns about whether diverse stakeholder perspectives—particularly from affected communities—are adequately represented in governance discussions.

AI governance remains fragmented across different sectors and regions: the reliance on specific frameworks like the EU AI Act and NIST RMF suggests there is no unified global approach, which could create compliance challenges for organizations operating internationally and inconsistent protections for AI users.