Quality of Authentic Assessment in Digital Marketing Education in the GenAI Era

Authentic assessment in digital marketing faces important challenges due to students’ limited access to real audiences and budgets, as well as their lack of time, while generative AI may create refined artefacts without making student reasoning visible. This paper presents a design-based research study of three assessment redesign cycles (2024–2026) in an undergraduate Digital Marketing course at the University of Patras, Greece. Three assignment briefs and 65 student submissions were analysed (2024 n = 19; 2025 n = 29; 2026 n = 17). Cycle 1 pursued authenticity of execution while Cycle 3 introduced a weighted rubric, market-gap evidence, benchmark expectations, and mandatory AI-use disclosure. A cross-cycle analysis distinguishes formal AI disclosure from Responsible AI competence: the 2026 artefacts made AI participation more visible but provided limited evidence of verification, risk awareness, and human accountability. The study offers transferable design principles for GenAI-era project assessment, namely explicit strategic simulation where real execution is unavailable, grading of justification chains, and embedded AI documentation, together with a cumulative model that distinguishes AI disclosure from Responsible AI competence for rubric design in applied business disciplines.