This study aims to examine how ethical concerns link to specific artificial intelligence (AI)-enabled marketing applications in public discourse and whether these linkages changed after ChatGPT’s public release on November 30, 2022. It addresses a research gap by analyzing AI marketing uses and AI ethics together, and tests whether a generative-AI shock reconfigures ethical attention across practice domains and stakeholder claim salience.
Using English language posts from X (2021–2023), the authors collected 584,941 tweets, retained 71,212 after quality filtering and analyzed 26,099 posts meeting an ethical concern threshold. Ethical concerns were defined from prior literature and assigned through semantic matching, while marketing applications were derived with BERTopic and explicit consolidation rules. The authors built weighted bipartite networks, compared pre/post-ChatGPT structures, identified communities and ran robustness checks for inference validity.
Results show a statistically supported structural reconfiguration after ChatGPT. Public discourse became more concentrated around AI-powered content creation, which emerged as the dominant application linked to multiple ethical concerns. Accuracy/misinformation and intellectual property concerns increased in centrality. Community patterns indicate redistributed ethical attention across operational, creative and growth-related marketing contexts, rather than a uniform rise across all applications in the network structure.
This study provides large-scale empirical evidence on how specific AI-enabled marketing uses connect to distinct ethical concerns in public discourse. Methodologically, it combines topic modeling, weighted bipartite network analysis and robustness tests to assess structural change. Theoretically, it advances ethics through diffusion of ethical risk and post-shock re-coupling. Managerially, the community map helps prioritize governance attention across marketing activities and stakeholder interfaces.
