Drawing on field research with Yezidi and Rohingya communities, this note examines ethical challenges of researching sexual violence in post-atrocity settings. Through these cases, I demonstrate how institutional ethics review processes can inadvertently reinforce power structures that harm vulnerable populations. The research reveals that choosing not to interview survivors directly can be both ethically necessary and methodologically sound, challenging assumptions about direct testimony’s centrality to rigorous research. Based on these insights, I propose reforms: mandating preliminary field assessments, requiring comprehensive evaluation of support services and emphasizing processual rather than procedural ethics. These findings have significant implications for transitional justice research and practice, particularly in truth-telling processes. They suggest moving beyond individualistic rights-based frameworks to incorporate indigenous approaches to justice and healing, while ensuring documentation efforts do not compromise community protection. These findings contribute to broader debates about ethical research practices, arguing for approaches that prioritize community protection over academic imperatives.