Confronting and managing ethical dilemmas in social work using ChatGPT

Abstract

Artificial intelligence (AI) is increasingly proving to be capable of assisting humans in complex decision-making processes. ChatGPT has become a crucial tool for assessing mental health and emotional awareness and helping to predict outcomes. Social workers’ use of ChatGPT in navigating ethical dilemmas remains relatively unexplored. In this study, ChatGPT is compared to the Israeli Ethics Committee’s responses to assess its utility in aiding social workers’ ethical decisions. We translated questions previously sent to the Israeli Ethics Committee and posed them to ChatGPT-4 twice, one month apart. A significant improvement was observed in ChatGPT's second round of responses, which were more comprehensive and emphasised ethical dilemma complexity. ChatGPT consistently highlighted conflicting obligations and professional boundaries, contributing to a more nuanced understanding of ethical issues. The Ethics Committee’s responses provided clearer, more actionable guidance, and was more concerned with legal issues. ChatGPT’s tendency to address a variety of considerations ensures its potential to enhance ethical decision-making awareness and its value as a tool for social workers but it is not yet capable of independently resolving ethical dilemmas. ChatGPT's promising progress in supporting ethical decision-making in social work has led to discussions regarding the ethical implications of AI implementation within therapeutic professions.