Governing artificial intelligence in public services: accountability, transparency and democratic AI governance

Purpose This paper aims to examine how democratic governments can regulate artificial intelligence (AI) in public services through accountable, transparent and institutionally governable frameworks. It addresses the governance gap between rapid AI adoption and limited democratic regulatory oversight by proposing the Transparent-Accountable-Governable (TAG) Framework. Design/methodology/approach This study adopts a qualitative conceptual research design using comparative policy analysis. It critically synthesises interdisciplinary literature in public administration, digital governance and AI ethics alongside the OECD AI Principles (2025), UNESCO Recommendation on the Ethics of Artificial Intelligence (2021), the National Institute of Standards and Technology (NIST) AI Risk Management Framework (2023) and the European Union Artificial Intelligence Act (2024) to develop the TAG Framework. Findings The analysis argues that trustworthy AI depends upon the interaction of transparency, accountability and governability rather than ethical principles alone. The proposed TAG Framework provides an institutional governance model applicable across welfare administration, healthcare, taxation, policing and disaster management. Research limitations/implications This is a conceptual study and therefore does not empirically validate the TAG Framework. Future research should evaluate its applicability through comparative cross-country case studies and institutional assessments of AI implementation within public administration. Practical implications The framework offers policymakers practical guidance by recommending public AI registers, algorithmic impact assessments, meaningful human oversight, independent auditing and adaptive regulatory coordination to strengthen democratic governance and institutional accountability. Social implications By emphasising explainability, citizen participation and procedural justice, the TAG Framework provides a conceptual basis for examining how AI governance can contribute to inclusive and equitable public services. It also highlights the risks of algorithmic bias and opaque decision-making, both of which may undermine public trust. Originality/value This paper proposes the TAG Framework to examine AI governance in democratic public administration. It brings transparency and accountability together with governability, understood as the capacity of public institutions to coordinate their activities, monitor AI systems, learn from experience and adapt governance arrangements over time.