The world is grappling with increasingly complex challenges, from global issues like climate change to local concerns such as rising asthma rates in communities near industrial zones. Tackling these problems requires collaboration between stakeholders across different domains, resulting in highly intricate, multi-stakeholder systems. These complex collaborations introduce unique challenges that make finding effective solutions difficult. While current design methodologies offer frameworks to support collaboration, their effectiveness diminishes as complexity grows, highlighting the urgent need for new, adaptive tools.

This research explores the potential of Large Language Model driven tools to enhance collaboration within these complex systems. It identifies key barriers to effective cooperation, including misaligned stakeholder values, communication breakdowns, and entrenched power dynamics. By addressing these challenges, LLM-powered tools offer a promising new approach to facilitate more inclusive, efficient, and adaptive collaborative processes. Through a combination of literature review and expert interviews, three critical themes that impact the succes of collaboration in complex systems were identified: Value Alignment, Communication & Certainty, and Power Structures. Traditional design methods are evaluated against these themes, highlighting their limitations in adequately addressing the complexities inherent in multi-stakeholder collaborations. To overcome these limitations, the study investigates the potential of LLM-based tools, notably leveraging OpenAI’s ChatGPT-4o model, to facilitate improved stakeholder interactions, streamline communication, and balance power dynamics.