Digital Twin of Legislation

Automated decision-making (ADM) in public administration must meet strict requirements of legality, transparency, and explainability. One potentially relevant but still underexplored route to achieve this is the Digital Twins of legislation, which can help synchronize legal texts, semantics, and executable decision logic. This paper investigates how such legislative digital twins can support explainable, rule-of-law–compliant ADM in administrative law. Based on grounded, inductive analysis of senior expert interviews, we identify four feasibility conditions for trustworthy ADM: (1) selective automation of deterministic sub-decisions, (2) semantic standardization and ontological alignment, (3) computable legal structures that link natural-language norms to machine-interpretable logic, and (4) drafting and governance adaptations enabling synchronized updates. Building on this, we derive design principles and a four-layer Digital Twin of Administrative Law (DTAL) architecture comprising statutory text, ontology, configuration, executable logic and illustrate its operation through a tourism contribution levy use case. This study contributes empirically grounded design principles, and a mid-range process theory explaining when and how administrative norms can be operationalized through digital twins. We propose an architectural approach that enables deterministic sub-decisions to be automated while preserving human oversight, legal traceability, and institutional legitimacy.

Administrative laws are traditionally published only in natural language, forcing each agency and business to manually encode identical rules into software, which is a redundant process that invites inconsistency and high maintenance costs. This paper introduces Digital Twins of Administrative Law (DTAL) as a centrally maintained, machine executable representation of legislation that remains synchronized with the authoritative legal text and preserves explicit traceability links. We propose a semi-automated, human in the loop pipeline that assists experts in transforming statutory provisions into an executable and testable DTAL bundle. The approach is evaluated in a real world case study on the Upper Austrian Tourism Contribution Levy using five implementation scenarios and a ground truth test set. Our results show that the centrally published DTAL can reduce administrative burden by an order of magnitude (35× reduction) and avoids duplicate law-to-code translations, while improving traceability and correctness of legal computations. Our work bridges legal informatics, public administration, and AI system design, and suggests policy implications for more transparent and efficient digital governance.

Public administrations increasingly explore automated decision-making, yet the normative content of legislation remains embedded in natural-language statutes rather than in machine-interpretable structures, so every agency, software vendor and business re-codes the same rules. This chapter introduces the Digital Twin of Legislation (DTL) – a synchronized, machine-executable representation of a defined set of legal provisions that is explicitly linked to the promulgated legal text and versioned alongside amendments – and operationalizes the vision of The Self-Driving State in five steps: it defines DTLs and derives design principles for rule-of-law-compatible automation, introduces a four-layer reference architecture linking legal text, semantics, parameters and executable logic, presents a semi-automated, human-in-the-loop pipeline for transforming statutes into DTL bundles, and evaluates the approach on the Upper Austrian tourism contribution case study, showing a reduction of collective implementation effort by approximately 35× under the study assumptions and full correctness on the expert-validated test set. Finally, it discusses governance implications for publishing, maintaining and certifying DTLs, emphasizing legal traceability and explainability as prerequisites for trust in AI-driven governance.

Automated decision-making in administrative law holds the promise of considerable efficiency gains, enhanced consistency, and increased transparency. This extended abstract distills a qualitative study based on semi-structured expert interviews that adopts a grounded theory approach to investigate the preconditions, constraints, and opportunities for embedding the concept of Digital Twin of Administrative Law (DTAL) into legislative and administrative processes while safeguarding the Rule of Law.

Automating public-sector decision-making promises efficiency gains in administration. This short paper proposes a research agenda for translating natural-language normative text into a Digital Twin of Administrative Law (DTAL), which we envision as a layered, executable representation of statutes that preserves traceability to the authoritative legal text while enabling accountable automation of decision-making in the public sector. To obtain a DTAL from normative text, a stepwise translation pipeline must be followed, which we demonstrated on the Upper Austrian Tourism Contribution Levy Act. The resulting DTAL yields deterministic and explainable outcomes. In this short paper, we outline open questions on standardizing legal ontologies, integrating LLMs as assistant tools, and embedding DTALs into public-sector engineering practices to realize transparent auditable automation of decision-making aligned with the rule of law.

Automated Decision-Making (ADM) in administrative law holds the promise of significant efficiency gains, enhanced consistency, and increased transparency. Yet, the inherent complexity of statutory drafting, semantic ambiguities, and evolving governance frameworks pose substantial challenges to its scalable implementation. This study adopts a Grounded Theory approach to investigate the preconditions, constraints, and opportunities for embedding Digital Twins of Administrative Law (DTAL) into legislative and administrative processes while safeguarding the Rule of Law. Drawing on semi-structured expert interviews with nine representatives from the European Parliament, academia, national parliament, ERP system vendors, and notarial practitioners, our findings identify four critical success factors for enabling ADM: (1) embracing only partial transformation, acknowledging that not all laws are automatable; (2) achieving semantic standardization and ontological alignment; (3) Transforming legislation into high-quality ADM-supporting structures; and (4) adapting legislative drafting processes and governance mechanisms to accommodate ADM. The results indicate that while partial automation is feasible, particularly in deterministic domains such as taxation and subsidies, most legal frameworks require hybrid models that preserve human judgment and interpretative flexibility.

Administrative-law provisions are still published almost exclusively in natural language, forcing every stakeholder to translate identical rules into bespoke code bases—a practice that invites inconsistency, hampers transparency, and inflates maintenance costs. Recent work on Digital Twins for Administrative Law (DTAL) suggests that legislation be issued together with machine-readable ontologies and executable logic, yet guidance on how to architect such systems remains scarce. In this work we propose a layered reference architecture that separates (i) the natural-language statute, (ii) a core ontology expressed in OWL, (iii) a configuration layer for mutable policy parameters, and (iv) an executable-rule layer exposed through a RESTful and MCP façade. Grounded in design science research, we implemented a proof-of-concept twin of the Upper-Austrian tourism-levy statute and qualitatively evaluate the twin with legal, software, and public-administration experts. Early results suggest that ontology-driven twins can reduce duplicate implementations, streamline updates, and enhance legal certainty, thereby strengthening the Rule of Law in automated decision-making.

Administrative laws are traditionally published in natural language, necessitating multiple, redundant software implementations that lead to inefficiencies and inconsistencies. In response, this early-stage Ph.D. research introduces a framework for co-publishing ontologies alongside statutory text, forming a digital twin for administrative law (DTAL). Using an Austrian tourism levy as a case study, this research investigates how simultaneously publishing ontology-based and legal prose representations can minimize redundant software development efforts, reduces errors, and support rule-of-law-compliant automated decision-making. Preliminary prototypes indicate measurable reductions in development effort and update latency, underscoring the approach's practical viability.

Administrative laws are traditionally published in natural language, necessitating multiple, redundant software implementations that lead to inefficiencies and inconsistencies. In response, this early stage Ph.D. research introduces a framework for co-publishing ontologies alongside statutory text, forming a digital twin for administrative law (DTAL). Using an Austrian tourism levy as a case study, this research investigates how simultaneously publishing ontology-based and legal prose representations can minimize redundant software development efforts, reduces errors, and support rule-of-law-compliant automated decision making. Preliminary prototypes indicate measurable reductions in development effort and update latency, underscoring the approach's practical viability. The poster condenses problem context, research question and methodology, state of the art and the four-layer DTAL architecture (text, ontology, model, logic) onto a single page.

Presentation deck of the doctoral research on Digital Twins for Administrative Law (DTAL). Laws are published exclusively in natural language, so every public and private organisation has to interpret them manually and re-implement identical rules in its own software – an inefficient, error-prone process that frequent amendments make worse and that puts the rule-of-law compliance of automated decision-making at risk. The talk asks which representation of administrative law enables automated decision-making under the rule of law while preserving the necessary interpretive flexibility, and proposes combining the structure of smart contracts, domain-specific legal ontologies and digital-twin principles from engineering into a four-layer architecture of text, ontology, model and logic that is implemented once and consumed by all systems via API. It outlines the design-science research plan with three sub-questions (semantic alignment, abstraction and complex dependencies, ADM with laws) and reports on the design cycles to date: the book “Der selbstfahrende Staat”, a prototype digital twin for parts of the Austrian income tax, a Prolog-based twin of the Upper Austrian landscape levy and the ongoing tourism-levy use case.

This paper presents proposed research on the design of a framework for automating decision-making in tax software systems through the development of digital twins of tax laws, using the Austrian Income Tax Act as an example. These digital counterparts of the legal texts formulated in natural language are structured into four layers, inspired by the structured approach of smart contracts used in automated legal contract systems: (1) the legal text, (2) a domain-specific tax ontology, (3) a model layer for data configurations, and (4) a logic layer where the tax law logic is implemented in a formal language. By formalizing these layers, we intend to create a comprehensive, machine-interpretable representation of (parts of) existing tax laws. Future work will further investigate which parts of law texts can be formalized in digital twins, which decisions can be automated, and where human inputs will continue to be required.