This chapter, initially drafted as and developed upon a policy paper for the DC-DAIG, examines the divergent conceptualizations of Artificial Intelligence (AI) accountability among diverse stakeholders and actors, laying the groundwork for a nuanced discourse on the inherent ambiguities and practical challenges in establishing normative frameworks for AI governance. In particular, the chapter critically assesses current global regulatory efforts targeting Generative AI, especially as they intersect with data protection legislation. It reveals how these varied regulatory approaches have coalesced into an interim, composite framework in the absence of comprehensive, formalized statutes. Using China as a focal jurisdictional case study, this analysis tracks the evolving regulatory landscape regarding ethical principles, content security, and data protection standards. The chapter juxtaposes the binding “Interim Measures for Generative AI” with the proposed, non-binding “Chinese Model AI Law,” introduced by leading Chinese academics in 2023 and iteratively revised in 2024, to illustrate contrasts and tensions in regulatory theory and practice. Ultimately, the chapter contends that achieving operational AI accountability demands a robust institutional framework that clearly defines specific obligations and liabilities, such as data retention, disclosure mandates, and mechanisms for cross-border cooperation. It further argues for a carefully calibrated balance between regulatory flexibility and predictability, essential to fostering an adaptable, enforceable, and practicable accountability framework for AI.