Regulating Artificial Intelligence in the Public Sector: Policy Frameworks for Accountability, Ethics, and Human Rights Protection

Abstract

The use of Artificial Intelligence (AI) in government institutions has changed the way government functions by improving its efficiency, decision making and services. But with growing dependence on AI systems, there have been serious concerns about accountability, ethical governance, and the safeguarding of fundamental human rights. The purpose of this study is to investigate policy frameworks that govern the use of AI in the public sector to discuss accountability mechanisms, examine the integration of ethics considerations, and determine human rights protection that is required for responsible AI governance. The study is informed by a rights-based approach to AI governance, which is based on theories in the broader area of accountability, ethical governance and human rights protection, to evaluate current governance frameworks. An Interpretivist research paradigm was employed and a qualitative research approach. Data collected through qualitative document analysis of 40 policy and regulatory documents (international conventions, AI regulations, governance guidelines, and national AI strategies), and 15 semi-structured interviews with policymakers, academics and civil society representatives. There were purposive sampling and snowball sampling techniques used in the selection of relevant documents and participants. Data collected were analysed using the six phases of the thematic analysis framework developed by Braun and Clarke with the help of the NVivo software. The results show that transparency, explainability, human supervision and independent auditing are the main mechanisms of accountability in current governance frameworks for AI. Fairness, transparency and non-discrimination were the most important ethical principles yet there are differences between the policy commitments and implementation. In addition, it was determined that privacy protection, equality safeguards, appeal mechanisms and independent oversight institutions are key human rights safeguards. Based on the findings of the study, it suggests that robust regulatory frameworks are essential to promote a society-centred and responsible use of AI in public services, incorporating accountability, ethics and human rights protections.