The gap between evidence discovery and using evidence-based decision-making is widened when those roles are filled by different groups. To bridge this gap and to more regularly translate evidence to practice we found that authentically including policymakers and other relevant stakeholders directly in the model building process, through events like policy modelling workshops, is a promising intervention we have coined as a trojan horse approach. Repeatedly, we have found that people are more likely to act on something if they discovered it themselves. Participatory co-modelling is thus a powerful intervention to reduce the discovery/usage gap. The 2024 Annual Modeling and Simulation Conference (ANNSIM) Workshop on Policy Modeling for Social Good, framed within the AI FORA US case study, brought together an interdisciplinary cohort of researchers, practitioners, and policymakers to explore how participatory modelling can inform public policy in domains such as healthcare, climate change, and social equity. This chapter synthesizes the workshop’s core themes, discussions, and takeaways, emphasizing methodological advances, institutional enablers and barriers, and ethical imperatives. This paper contributes to the growing literature on computational social science and evidence-based decision-making. The integration of modelling and simulation (M&S) into policymaking presents transformative potential for addressing complex social challenges and responding to complex demands by various stakeholders.