Social Work Ethics in the Age of Algorithmic Decision-Making: Rethinking Confidentiality and Informed Consent

Abstract: Algorithms are entering child protection and welfare casework faster than the ethical frameworks that govern social work practice can accommodate them. This paper is not concerned with whether such tools are effective; it is concerned with whether confidentiality, informed consent, and professional accountability continue to function once decision-making authority is distributed between a caseworker and a computational system.

We ground the argument in Floridi's information ethics, read alongside the ethics of care and the data justice literature, to show that algorithmic mediation destabilises these commitments at a structural rather than incidental level. Drawing on critical policy analysis of the NASW Code of Ethics, the IFSW Global Statement of Ethical Principles, a systematically identified body of recent empirical and conceptual scholarship on algorithmic decision-making across child welfare, public administration, and healthcare, and current regulatory instruments — the EU AI Act as amended by the 2026 Digital Omnibus, the OECD AI Principles, and India's newly notified Digital Personal Data Protection Rules, 2025 — we identify three structural gaps. Consent frameworks were not designed to anticipate downstream, predictive reuse of client data. Confidentiality provisions presume a practitioner control over information that has, in practice, migrated into opaque proprietary systems. Accountability mechanisms remain fixed on the individual practitioner even as decisional authority has shifted toward institutional and commercial actors.

We examine the counter-argument that algorithmic tools may reduce inconsistency and bias in casework, and we show why this claim, even where empirically supported, resolves only one of the three gaps and leaves consent and accountability untouched. The paper closes with an original accountability framework organised around layered and renewable consent, mandated algorithmic transparency toward practitioners, and institutional co-responsibility for automated decisions, with attention to its implications for social work practice, professional education, and policy.

Keywords: social work ethics, algorithmic decision-making, informed consent, confidentiality, information ethics, data justice, digital welfare