Generative Artificial Intelligence in Graphic Design Higher Education: A Comparative Integrative Review of Western and Asian Evidence and the CACG Framework

Abstract

Generative artificial intelligence (GenAI) has moved from an optional image-making technology to a contested component of graphic design education. Yet the evidence available to curriculum leaders remains fragmented across graphic design, visual arts, design engineering, educational technology, assessment, and policy research. This article presents a comparative integrative review of scholarship published from January 2019 to 20 August 2026, with emphasis on differences and overlaps between selected Western higher-education systems and Asian contexts. The analytic corpus comprised 39 peer-reviewed publications and seven authoritative policy or regulatory documents identified through structured searches of scholarly publisher platforms, DOI-verified records, and backward and forward citation chaining. Evidence was coded across curriculum, assessment, capability, and governance. The review finds consistent short-term benefits for ideation speed, visualisation, self-efficacy, and exploratory breadth, but substantially weaker evidence for durable gains in disciplinary knowledge, craft, independent judgement, or employability. Western studies more often foreground authorship, critique, assessment validity, and the material politics of AI, whereas recent Asian studies more often employ acceptance, self-efficacy, visual-literacy, and institutional-readiness models. These are tendencies rather than stable cultural differences, and the contrast partly reflects research design, policy maturity, and uneven access to infrastructure. Across regions, the central pedagogical risk is not simply plagiarism; it is the substitution of plausible output for observable learning. To address this problem, the article proposes the Curriculum-Assessment-Capability-Governance (CACG) framework for accountable human-AI design learning. The framework aligns AI visual literacy and foundational design knowledge with process-visible assessment, staff and student capability, and institution-level rules for provenance, copyright, data, access, and platform review. For Malaysia and Southeast Asia, the immediate research priority is discipline-specific, multi-institutional evidence that measures creative processes and learning outcomes rather than adoption intention alone. Keywords: generative artificial intelligence; graphic design education; visual communication; higher education; artificial intelligence literacy; assessment; comparative review; Malaysia; CACG framework