Ethical Implications of AI‐Driven Chatbots in Domestic Violence Support

Abstract

Our study explored the opportunities, challenges, and ethical considerations of using artificial intelligence (AI)‐driven chatbots in domestic violence (DV) support. DV is a serious public health and social problem. Identifying it as early as possible is important in violence prevention. However, victim‐survivors of DV are often reluctant to disclose violence, and service practitioners may lack the capacity or confidence to address violence‐related issues. To tackle these challenges, the use of AI‐driven chatbots presents opportunities to address DV by providing information and guiding users to appropriate services. However, interactions between humans and AI systems lie at the intersection of the human need for practical assistance and the risks inherent in digital communication - raising ethical considerations, particularly in the vulnerable context of DV. Semi‐structured interviews with 25 victim‐survivors, DV professionals, and criminal justice experts suggest that while the implementation of AI‐driven chatbots can greatly enhance access to information, it also poses significant challenges related to safety and accountability. This is because interactions with chatbots lack essential elements for comprehensive situational assessment and documentation of DV cases, and for the establishment of a support network. These insights underscore the critical role of human interaction in addressing DV cases, while also highlighting the potential roles of chatbots as intermediate support systems for victim‐survivors and as supplementary tools for welfare service practitioners in identifying different forms of DV and supporting the appropriate referral of cases. This study advances understanding of how AI‐driven chatbots can be ethically and sustainably implemented in DV support systems.