Governance of Generative Artificial Intelligence for Companies
While generative AI (GenAI) systems—such as ChatGPT—are being rapidly deployed in enterprises, existing AI governance frameworks fail to address GenAI’s unique technical characteristics (e.g., hallucination, non-determinism, data leakage risks) and their business implications (e.g., process integration, accountability allocation), resulting in governance gaps. Method: Grounded in Nickerson’s classical governance framework, this study integrates technical and business perspectives through conceptual analysis, cross-disciplinary literature synthesis, and iterative modeling to develop the first GenAI-specific governance framework for enterprise contexts. Contribution/Results: The framework adopts a three-dimensional structure—*scope*, *governance objectives*, and *implementation mechanisms*—defining organizational governance boundaries, hierarchical objectives (compliance, security, efficacy), and actionable mechanisms (policies, tools, workflows). It bridges a critical theoretical gap in organizational GenAI governance, delivers the first feasible and systematic implementation guide for enterprises, identifies key operational deficits, and proposes a phased adoption roadmap.