This study systematically reviewed the existing literature to identify social work intervention areas where Artificial Intelligence (AI) could be most useful and the ethical issues that may arise as a result of the use of AI in practice. AI and Information and Communication Technology (ICT) have recently seen use and have reshaped social work practice. Technology-based interventions could contribute to social work practice in many ways, such as predicting and preventing crisis response and health issues, improving delivery systems, and better predictions in decision- making. Also, AI or ICT could help marginalized populations by age, disability, sexual orientation, geography, or economics (Berzin et al., 2015). Compared to the other disciplines, however, social work has been slow to adopt AI and is still in its infancy due to limited resources, ethical concerns, lack of training and social work’s reliance on direct, face-to-face communications (Mishna et al., 2015).
Utilizing the PRISMA guidelines, this study conducted a systematic review of the literature. I employed electronic databases, hand searching of electronic journals, and citation tracking. I searched the following electronic databases: APA PsycInfo; MEDLINE; Nursing and Allied Health; Social Service Abstracts; and Social Work Abstracts. Searches consisted of a combination of keywords that included: “Artificial Intelligence” OR “Information and Communication Technology” OR “Machine Learning” OR “Robotics,” paired with AND “social work practice” OR “social work” OR “social care” OR “social good.” They were also combined with “ethics” OR “risks.” Inclusion criteria for searches were studies written in English published between 2000 and 2020.
A comprehensive literature review of this study finds that the social work intervention areas where AI could be most useful include: child protective services, health and mental health services, elderly care, and sexual minority youth. Regarding ethical considerations, this study discusses the following: a loss of privacy and confidentiality resulting from the monitoring system, the potential reduction of human contact and quality of care, people’s loss of dignity, and encouragement of dependency. Also, concerns about AI systems are reinforcing inequality and existing biases to gender and race. Furthermore, the literature tells social workers’ job loss and loss of autonomy.
Conclusions and Implications:
The findings of the study offer some implications for using AI in social work practice. The use of AI in social work interventions has the great potential to serve people who are previously marginalized or whose problems do not fit within the traditional social work framework. Evidence on the use of AI in social work interventions is still limited. Also, a lack of empirical research on AI and technology-based interventions has inhibited social work practitioners’ AI application to their practice. This study suggests that social work requires more in-depth research to use AI ethically and effectively. Although there are ethical concerns in applying AI into social work intervention, there are potentially far‐reaching benefits. Also, there is a need for interdisciplinary collaboration with other disciplines, such as computer science, engineering, and businesses, to help social work effectively adopt AI.