Salvatore Stanizzi (- Department of Finance; Magna Graecia University of Catanzaro – Department of Legal, Historical, Economic and Social Science, Students) has posted From Turing to Algorithmic Governance: The Legal Limits of Computability on SSRN. Here is the abstract:
This paper argues that contemporary debates on algorithmic governance rest on an unexamined assumption: that public decision-making can be fully formalized and optimized through computation. By revisiting Alan Turing’s theory of computability and the distinction between decidable and undecidable problems, the paper contends that legal reasoning cannot be entirely reduced to executable instructions without incurring normative loss. Law is not merely a rule-based system but an interpretative and context-sensitive practice marked by structural indeterminacy. Integrating insights from legal theory and law & economics, the paper shows that while algorithmic systems may reduce decision and information costs, they generate a category of non-computable costs-including legitimacy deficits, accountability gaps, and adaptive rigidity-that cannot be internalized within computational optimization frameworks. These costs emerge precisely from the mismatch between formalizable processes and the open-textured nature of legal judgment. The analysis concludes that algorithmic governance must incorporate the limits of computability as a structural design principle. Rather than pursuing total automation, legal systems should recognize irreducible domains of human judgment and contestability. The central risk of algorithmic governance, therefore, is not technical failure but normative overconfidence: the belief that computability exhausts legality.