This study contributes to the development of ecoantitrust, an approach to antitrust regulation that draws on analogies from biology and ecology to analyze and monitor the behavior of major technology companies and digital ecosystems. It examines how expert engagement can support the development of systems mapping for competition analysis, using digital platforms developed by major agricultural commodity traders as a case study. Drawing on two expert workshops held in 2025, we investigate the opportunities and challenges of eliciting competition-relevant knowledge and translating it into systems maps using Large Language Models (LLMs). The authors construct and compare six candidate systems maps based on workshop records, consolidated lists of factors, and selected literature. The workshops proved effective in identifying potential competition concerns, benefits, and responses, but less successful in eliciting explicit causal relationships. AI-assisted analysis helped recover relationships not captured in the consolidated lists and revealed complementary contributions from expert discussions and literature. However, the resulting maps, particularly their feedback structures, were rather sensitive to methodological choices. The findings demonstrate the potential of combining expert engagement with AI-assisted systems mapping while highlighting the importance of careful elicitation design, transparent processing, and expert validation.
The International Institute for Applied Systems Analysis (IIASA)