Legal Accountability in Algorithmic Public Decision Making and the Transformation of Administrative Governance

The article begins by explaining that artificial intelligence and algorithmic systems are increasingly used in public decision-making. Governments rely on automated or semi-automated tools in welfare distribution, immigration screening, predictive policing, tax enforcement, public health administration, and judicial risk assessment. This development is linked to digital government, data-driven regulation, and administrative efficiency.

Algorithmic governance is often presented as a rational response to the complexity and speed of modern public administration. Public institutions are expected to process large quantities of data, manage limited resources, and provide services in a consistent and timely way. Algorithmic systems therefore appear to offer precision, efficiency, and objectivity in public decision-making.

The article stresses that public decisions are not merely technical outputs because they involve exercises of state power. Decisions affecting liberty, livelihood, mobility, social protection, or legal status must remain subject to legality, accountability, and procedural fairness. When such decisions are shaped by opaque computational models, the relationship between individuals and the state becomes harder to scrutinize.

The article identifies key risks in algorithmic public decision-making. Individuals may not receive meaningful explanations, public officials may defer too strongly to algorithmic recommendations, and institutions may find it difficult to assign legal responsibility when harm occurs. Algorithmic systems may also reproduce social bias, institutional inequality, and structural discrimination at scale.

Existing scholarship has examined algorithmic governance through ethics, regulation, public administration, transparency, data protection, administrative review, and human oversight. These studies show that algorithmic tools are not neutral technical instruments, but are embedded in political and organizational structures that can transform public administration and its legitimacy.

However, the article argues that current scholarship still lacks conceptual precision regarding legal accountability under algorithmic governance. Many discussions treat accountability as a broad aspiration, or focus on isolated principles such as transparency and explainability. Less attention is given to how these principles interact within the wider structure of administrative legality.

The article identifies a research gap at the intersection of legal theory and administrative law. Traditional administrative law assumes that decisions can be attributed to identifiable authorities, justified through accessible reasons, and reviewed through established mechanisms. Algorithmic governance disrupts these assumptions because decision pathways may be hidden, technically complex, and organizationally fragmented.

The article aims to clarify how legal accountability should be reconstructed when public authority is technologically mediated. It asks how algorithmic governance transforms accountability, which traditional principles remain applicable, how answerability and justification should function, and whether public law can preserve meaningful control over algorithmically mediated authority.

The article uses a qualitative doctrinal and socio-legal research design. This approach is suitable because the study examines normative meaning, institutional logic, and conceptual tensions rather than statistical relationships or causal measurement. The analysis combines legal accountability theory, administrative law, and governance studies to examine how answerability, justification, reviewability, and responsibility operate when public decisions are shaped by algorithmic systems.

The data sources include primary legal materials and secondary interpretive materials. Primary sources consist of statutes, administrative regulations, judicial decisions, policy guidelines, and official governance documents related to algorithmic decision-making and public administration. Secondary sources include scholarly articles, legal commentaries, institutional reports, and interdisciplinary literature on artificial intelligence, governance, and accountability. The materials are analyzed through dimensions such as transparency, explainability, human oversight, reviewability, contestability, and allocation of legal responsibility.

The article finds that algorithmic governance changes the legal structure of public decision-making by relocating parts of administrative judgment into data infrastructures, computational models, and automated recommendation systems. Public authority is no longer exercised only through visible human interpretation, but increasingly through hybrid arrangements involving legal discretion and technical architectures.

Conventional public law principles remain important in algorithmic governance. Legality, procedural fairness, reason-giving, and reviewability still apply because algorithmic systems do not remove the constitutional and administrative obligations attached to public decision-making. However, these principles must operate under new institutional conditions shaped by automated systems.

Legality must now address not only whether a decision is formally authorized, but also whether the design and deployment of decision-support systems are lawful. Procedural fairness must include the ability of affected individuals to understand how automated processing shapes official judgment. Reviewability must also respond to the fact that reasons may be embedded in technical processes that ordinary legal scrutiny cannot easily access.

The article argues that transparency alone is insufficient to sustain legal accountability. Disclosure of source code, technical documentation, or general system descriptions does not automatically provide reasons that affected individuals can understand or challenge. Accountability requires more than visibility; it requires intelligible explanations connected to legal standards.

A key distinction is made between visibility and intelligibility. A system can be procedurally visible but still normatively obscure if its operational logic cannot be translated into legally meaningful reasons. Public law accountability therefore depends on explanations that connect institutional action to justification, proportionality, and lawful authority.

Responsibility attribution becomes unstable in algorithmic governance. Automated systems may be developed by private vendors, trained on datasets from multiple administrative sources, and implemented by public bodies with limited technical control over system design. This creates uncertainty over who is legally answerable when harm occurs.

The article shows that human oversight is not automatically sufficient. Human involvement is often presented as a safeguard, but it matters only if it involves genuine evaluation. If officials simply defer to algorithmic outputs because they are faster, normalized, or technically authoritative, oversight becomes formal rather than substantive.

The legal meaning of human oversight must therefore be reconstructed around active judgment, reasoned departure, and institutional willingness to question automated recommendations. A formally human decision may still reproduce the authority of the algorithm if the official does not meaningfully assess the output.

The article organizes legal accountability through dimensions such as answerability, justification, transparency, human oversight, reviewability, responsibility attribution, and contestability. These dimensions are interdependent. Weakness in one area affects the others, because effective challenge depends on explanation, reviewability depends on traceability, and responsibility depends on clear attribution.

Algorithmic governance therefore destabilizes accountability across the full chain of public decision-making. The problem is not only opacity, but the reorganization of the legal environment in which decisions are exercised, scrutinized, and remedied. Accountability must be treated as a composite legal structure rather than a single procedural safeguard.

The article argues that legal accountability should extend across the lifecycle of algorithmic systems. Relevant legal concerns include datasets, model design, procurement practices, implementation protocols, operational oversight, and review mechanisms. Responsibility may be shared and layered, but it must not become indeterminate.

The article concludes that public law remains capable of regulating algorithmic authority if it expands its analytical vocabulary. Public institutions need accountability frameworks that integrate system design, operational oversight, legal review, evidentiary access, institutional traceability, and meaningful contestability. Trust in digital government depends on durable legal answerability rather than innovation rhetoric alone.

Algorithmic governance is reshaping the legal foundations of public decision-making by embedding administrative judgment within technical systems, data infrastructures, and institutional arrangements that complicate answerability, justification, and review. The discussion has established that conventional public law principles remain normatively relevant, yet their practical operation is increasingly strained when decision pathways become opaque, responsibility is dispersed, and human oversight functions only formally rather than substantively. Transparency alone is insufficient to preserve accountability because disclosure does not necessarily produce legally meaningful explanations or effective opportunities for challenge. The central issue lies in the transformation of accountability from a relatively stable administrative relationship into a layered legal structure that must now address technical mediation, fragmented authority, and reduced institutional traceability. Public decision-making in digital environments therefore requires a more precise legal understanding of how responsibility is assigned, how reasons are communicated, and how review remains possible under conditions of algorithmic complexity.

The contribution to legal scholarship lies in the reconstruction of legal accountability as a composite framework for evaluating algorithmically mediated public authority. Rather than treating accountability as a narrow procedural safeguard or a general normative aspiration, the analysis has clarified its interdependent dimensions, including answerability, justification, transparency, reviewability, contestability, and responsibility attribution. This conceptualization strengthens the dialogue between administrative law, legal theory, and governance studies by showing that the primary challenge of algorithmic governance is not merely technological opacity but the destabilization of the legal form through which public authority is rendered intelligible and contestable. The argument also advances current debates by moving beyond compliance-based approaches and highlighting the doctrinal significance of upstream processes such as system design, procurement, and institutional implementation. In this way, the discussion reinforces the relevance of public law as a normative framework capable of engaging digital transformation without surrendering its core concern with legality, institutional responsibility, and protection against arbitrary power.

Future research should expand this line of inquiry through comparative, sector-specific, and interdisciplinary analysis. Comparative work across jurisdictions would be particularly valuable for examining how constitutional traditions, regulatory models, and administrative cultures influence the structure of accountability in algorithmic governance. More focused studies on welfare administration, immigration control, policing, taxation, and judicial systems would further clarify how accountability pressures vary according to the legal stakes, institutional design, and degree of automation involved in different domains of state action. There is also a need for deeper investigation into the relationship between legal doctrine and digital infrastructure, especially regarding evidentiary access, system traceability, and the design of review mechanisms capable of responding to technically mediated decisions. Such research would support the development of more robust legal standards that preserve meaningful answerability and contestability as public administration becomes increasingly dependent on algorithmic systems.

Adhabi, E., & Anozie, C. L. (2017). Literature Review for the Type of Interview in Qualitative Research. International Journal of Education, 9, 86–97. https://doi.org/10.5296/ije.v9i3.11483

Ananny, M., & Crawford, K. (2018). Seeing without knowing: Limitations of the transparency ideal and its application to algorithmic accountability. New Media & Society, 20, 973–989. https://doi.org/10.1177/1461444816676645

Bannister, F., & Connolly, R. (2020). Administration by algorithm: A risk management framework. Information Polity, 25, 471–490. https://doi.org/10.3233/ip-200249

Cheong, B. C. (2024). Transparency and accountability in AI systems: safeguarding wellbeing in the age of algorithmic decision-making. Frontiers in Human Dynamics. https://doi.org/10.3389/fhumd.2024.1421273

Coeckelbergh, M. (2019). Artificial Intelligence, Responsibility Attribution, and a Relational Justification of Explainability. Science and Engineering Ethics, 26, 2051–2068. https://doi.org/10.1007/s11948-019-00146-8

Edelmann, A., Wolff, T., Montagne, D., & Bail, C. (2020). Computational Social Science and Sociology. Annual Review of Sociology, 46 1, 61–81. https://doi.org/10.1146/annurev-soc-121919-054621

Enarsson, T., Enqvist, L., & Naarttijärvi, M. (2021). Approaching the human in the loop – legal perspectives on hybrid human/algorithmic decision-making in three contexts. Information & Communications Technology Law, 31, 123–153. https://doi.org/10.1080/13600834.2021.1958860

Engelmann, A. (2023). Algorithmic transparency as a fundamental right in the democratic rule of law. Brazilian Journal of Law, Technology and Innovation. https://doi.org/10.59224/bjlti.v1i2.169-188

Goldsmith, S., & Yang, J. (2025). AI and the Transformation of Accountability and Discretion in Urban Governance. ArXiv, abs/2502.13101. https://doi.org/10.48550/arxiv.2502.13101

Horneber, D., & Laumer, S. (2023). Algorithmic Accountability. Business & Information Systems Engineering, 65, 723–730. https://doi.org/10.1007/s12599-023-00817-8

Khasawneh, F., Amoush, M. A. M., & Khwaileh, K. (2025). Governing Algorithms: The Intellectualization of Administrative Law and Emerging International-Law Constraints. Veredas Do Direito. https://doi.org/10.18623/rvd.v22.n3.3464

Kurniawan, M. J., & Purwanto, R. (2025). The Revolution of Public Administration in the Era of Artificial Intelligence: Legal Challenges and Constitutional Implications. Fortiori Law Journal. https://doi.org/10.47200/flj.v5i01.3050

Kuziemski, M., & Misuraca, G. (2020). AI governance in the public sector: Three tales from the frontiers of automated decision-making in democratic settings. Telecommunications Policy, 44, 101976. https://doi.org/10.1016/j.telpol.2020.101976

Lepri, B., Oliver, N., Letouzé, E., Pentland, A., & Vinck, P. (2017). Fair, Transparent, and Accountable Algorithmic Decision-making Processes. Philosophy & Technology, 31, 611–627. https://doi.org/10.1007/s13347-017-0279-x

Liu, H.-W., Lin, C.-F., & Chen, Y.-J. (2019). Beyond State v Loomis: artificial intelligence, government algorithmization and accountability. International Journal of Law and Information Technology, 27, 122–141. https://doi.org/10.1093/ijlit/eaz001

Mark, D., & Morison, J. (2025). Governing AI Decision-Making: Balancing Innovation and Accountability. Politics and Governance. https://doi.org/10.17645/pag.10245

McGregor, L., Murray, D., & Ng, V. (2019). International Human Rights Law as a Framework for Algorithmic Accountability. International and Comparative Law Quarterly, 68, 309–343. https://doi.org/10.1017/s0020589319000046

Moch, E. (2025). Algorithms as Judges: Constitutional Limits to Automated Decision-Making in Public Administration. East African Journal of Law and Ethics. https://doi.org/10.37284/eajle.8.1.3044

Moon, K., Brewer, T., Januchowski-Hartley, S., Adams, V., & Blackman, D. (2016). A guideline to improve qualitative social science publishing in ecology and conservation journals. Ecology and Society, 21, 1–20. https://doi.org/10.5751/es-08663-210317

Reidenberg, J., Reidenberg, J., Robinson, D., & YUt, H. (2016). Accountable Algorithms. https://consensus.app/papers/accountable-algorithms-reidenberg-reidenberg/246a3e3340f65b8d8824062c91ec655b/

Rumlus, M. H. (2025). Governing AI: A Narrative Review of Algorithmic Accountability and Legal Frameworks. Legalis: Journal of Law Review. https://doi.org/10.61978/legalis.v3i2.810

Sharma, S. (2025). Governance in the Age of Algorithms: Ethical Dilemmas and Administrative Reforms. International Journal of English Literature and Social Sciences. https://doi.org/10.22161/ijels.102.58

Silfia, R. I. (2025). Role of Administrative Law in Regulating Advanced Algorithms: A Legal Framework for Emerging Technologies. Best Journal of Administration and Management. https://doi.org/10.56403/bejam.v3i4.278

Waldman, A., & Martin, K. (2022). Governing algorithmic decisions: The role of decision importance and governance on perceived legitimacy of algorithmic decisions. Big Data & Society, 9. https://doi.org/10.1177/20539517221100449

Wieringa, M. (2020). What to account for when accounting for algorithms: a systematic literature review on algorithmic accountability. Proceedings of the 2020 Conference on Fairness, Accountability, and Transparency. https://doi.org/10.1145/3351095.3372833