Back to Search View Original Cite This Article

Abstract

<jats:p>This study examines how overdevelopment and its associated ecological harm are represented in outputs generated through generative artificial intelligence (Gen AI) systems. The analysis draws on 96 outputs from ChatGPT, Claude, Gemini, and DeepSeek, coded at the meaning-unit level using a comparative qualitative approach and a hybrid deductive-inductive framework. The findings show that representations varied across systems and prompt conditions. Under consequence-oriented prompts, the outputs focused mainly on habitat loss, biodiversity decline, and ecosystem degradation. Under causal prompts, speculative development pressures, regulatory weaknesses, and construction-driven economic growth became more visible. Actor visibility was also uneven. The Planning Authority and environmental NGOs appeared relatively frequently, while scientists and experts and EU institutions showed the clearest prompt-specific gaps. The study shows that differences across systems and prompt conditions concern not only the information provided, but also which actors, explanations, and policy responses become visible. It therefore approaches Gen AI systems as components of epistemic mediation processes and conceptualises epistemic inequality as uneven representation of the same socio-ecological issue.</jats:p>

Show More

Keywords

systems outputs study prompt conditions

Related Articles

PORE

About

Connect