This is the authors' abstract. We don't add key points for this paper.
Balancing creative originality with strategic discipline remains a fundamental paradox in advertising creativity. Although generative artificial intelligence (GenAI) can produce novel recombinations of creative elements, it may not reliably align with brand strategy and campaign-specific executional requirements. To address this tension, organizations are reconfiguring creative workflows through new forms of human-AI collaboration. Drawing on an inductive multi-case study of five advertising organizations, including an incumbent agency and an AI-native startup, this study develops the 4A Modes Framework of Human-AI Collaboration. Organized along two dimensions of AI autonomy and process embeddedness, this framework identifies four modes of collaboration: Augmentation, Automation, Agentic, and Alliance. Furthermore, we identify two evolutionary paths through which organizations systematically navigate this tension: Process Integration and Native Innovation. Across both paths, micro-ambidexterity enables practitioners and teams to flexibly switch between and recombine collaboration modes within a single production process. Overall, these findings provide a theoretical lens for understanding the structural transformation of creative production and offer practical guidance for achieving effective and sustainable human-AI symbiosis.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Bao et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: