Preserving the Integrity of Evidence-Based Social Work in the Age of AI: A Proposed Ethical Framework

The purpose of this introductory article is three-fold: (1) to share with the reader the inspiration for the present special issue; (2) to describe for the reader the array of articles published in this special issue and the organizing logic; and (3) to offer for the readers’ consideration a proposed conceptual framework for understanding the ethical role of Artificial Intelligence in the social work profession. In 2024, David Edmonds and a team of leading philosophers published AI Morality. Edmonds identified six egent themes based on the authored chapters: autonomy; bias; responsibility; privacy and transparency; meaning; and, values and morals. We explore these six themes as a viable complement to the Belmont Report and the National Association of Social Workers (NASW) Code of Ethics for understanding categorical ethical concerns related to the use of AI in social work research, and in this way, evidence-based social work practice and broader society. We thematically grouped the corresponding relationships and used these groupings as an organizing framework: (1) autonomy, power/oppression, and informed consent; (2) bias, discrimination, and social justice; (3) responsibility, harm, and competence; (4) privacy, confidentiality, and transparency; (5) meaning, service, and social need; and, (6) values, morality, and ethical alignment. We identify ethical concerns across these six categories and make respective recommendations before offering final thoughts.