The rapid diffusion of generative artificial intelligence (AI) has created new challenges for course design and assessment in business education, particularly where instructors must balance innovation, academic integrity, and evaluative clarity. While much of the emerging literature focuses on student outcomes or ethical evaluation, less attention has been given to how ethical AI use is operationalized through course design and assessment architecture. This study presents a descriptive, design-focused analysis of how generative AI policies, instructional constraints, and assessment criteria were embedded across multiple undergraduate business courses. Drawing on extant instructional artifacts—including syllabi, assignment prompts, weekly instructional scaffolding, and grading rubrics—the analysis maps distinct design approaches to AI integration and identifies recurring patterns in assessment alignment. The findings highlight how transparency-by-design, bounded AI permissions, and rubric-level articulation of expectations function as governance mechanisms within routine instructional practice. By focusing on course architecture rather than student outcomes, this study contributes a practical framework for instructors seeking to integrate generative AI into business education while preserving assessment integrity and pedagogical coherence.