Humanitarian organizations are increasingly adopting artificial intelligence (AI) technology for their programming and operations. The rapid uptake of AI technologies has been a source of active debate in the humanitarian sector, including discussion of the best ways to harness its potential while avoiding negative impacts on humanitarian operations. Amongst these concerns are questions related to security and confidentiality, potential bias and discrimination, and a lack of accountability in the use of new AI technology. These issues highlight the need for humanitarian agencies to address the ethical implications of AI adoption, particularly in terms of its long-lasting effects on crisis-affected groups and individuals. The short-term emergency nature of humanitarian aid entails that humanitarian projects will eventually close once the crisis subsides, project objectives are met or due to budgetary and security constraints. In this chapter, we examine the ethical issues associated with using AI in the closure of humanitarian projects. We begin by outlining the context of humanitarian project closure and reviewing the current uses of AI in project closure and their ethical implications. We then build on the ethical capacities proposed by Pal et al. (2019)—attentiveness, responsiveness, and foresighting—to explore how these capacities can support the responsible implementation of AI in project closure and serve as a means to assess its risks and opportunities. Additionally, we discuss the importance of a strong organizational ethical infrastructure, including ethical resources, practices, and expertise, to support the enactment of attentiveness, responsiveness and foresighting related to AI during project closure. We describe how the fostering and practice of these ethical capacities can support humanitarian organizations to utilize AI in their operations while upholding their commitments to promote the well-being of affected communities during and after project closure.