Adopting AI Enhances Humanitarian Operations While Demanding Critical Trade-Offs

Abstract Artificial intelligence (AI) is increasingly impacting the cooperation sector, transforming how humanitarians are operating in the field. AI tools now allow for improved health diagnostics and quality services. In conflict areas, organizations are able to improve the efficiency of their analysis, and propose optimized data management processes for more multifactorial predictions. In a context where human and financial resources are limited, AI address the gaps to sustain emergency responses. In this commentary, we explore the trade-offs of adopting those tools and the ethical concerns their usages raise in a fragile environment. Security breach due to human errors and unregulated data management are a substantial risk to the humanitarian workers but also to the communities they aim to protect via their interventions. In this work, we touch upon humanitarians' concerns with this new technology, and reflect on balancing those with the benefits of accuracy of interventions. We argue that the humanitarian sector must not overlook these technologies, as failing to adopt them, risks widening inequalities in access to data-driven decision-making and leaving vulnerable systems further behind. At the same time, it is essential to establish meaningful safeguards and promote local ownership to ensure that AI enhances resilience, rather than reinforcing existing power imbalances.