Jusletter IT

«Elektronische Rechtsetzung» steht zum einen für die informationstechnische Unterstützung von Rechtsetzungsprozessen – also von hochformalisierten Geschäftsprozessen, die durch verfassungs- und verfahrensrechtliche Regeln gesteuert werden, welche eine tief in die Papierkultur zurückreichende Tradition aufweisen –, zum anderen für neue Umgebungen, in denen sich Rechtsetzung vollzieht, und zwar nicht nur technische, sondern auch soziale Umgebungen, die durch die neuen Informations- und Kommunikationstechnologien geprägt sind. Österreich nimmt in der elektronischen Rechtsetzung im europäischen Rahmen eine Vorreiterrolle ein; umso mehr erscheint es geboten, die jeweils aktuellen Entwicklungen zu beobachten und zur Grundlage für den weiteren Ausbau legistikspezifischer Anwendungen zu machen. Aus diesem Grund richtet der Workshop «Elektronische Rechtsetzung» den Blick über die Grenzen – über die Grenzen Österreichs, vor allem aber auch über die Grenzen der heute operativen Systeme.

Der Perspektive auf eine weiter wachsende Bedeutung der neuen Technologien für die Unterstützung der Dynamik des normativen Systems entsprechend, widmet sich bereits zum dreizehnten Mal im Rahmen des IRIS ein Schwerpunkt dem Thema «Elektronische Rechtsetzung».

Der erste Workshop dieser Art, im Jahr 2008, hat das Hauptaugenmerk auf elektronische Anwendungen zur Unterstützung zivilgesellschaftlicher Partizipation im Rechtsetzungsprozess gerichtet. Im Mittelpunkt des zweiten Workshop, im Jahr 2009, sind die neuen Ansätze des Wissensmanagement und insbesondere der semantischen Technologien in ihrer Anwendbarkeit auf den Rechtsetzungsprozess gestanden. Der dritte Workshop, im Jahr 2010, hat sich besonders mit der Frage auseinandergesetzt, welche Anforderungen an die Gestaltung legistischer Arbeitsumgebungen zu richten sind und wie solche Umgebungen künftig beschaffen sein könnten. Im Rahmen des vierten Workshop, im Jahr 2011, ist die Beschäftigung mit diesem Thema fortgeführt, der Schwerpunkt aber auf die Diskussion der Möglichkeiten der semantischen Modellierung für Rechtsetzung und Rechtsfolgenabschätzung gelegt worden. Der fünfte Workshop hat sich entsprechend dem Generalthema, unter welchem IRIS 2012 gestanden ist, nämlich «Transformation juristischer Sprachen», schwerpunktmäßig mit der Sprache der Rechtsetzung befasst. Der sechste Workshop hat sich der zeitgemäßen Gestaltung legislativer Konsultationsprozesse sowie der Entwicklung semantischer Werkzeuge zur Substituierung oder Ergänzung solcher Prozesse durch automationsunterstütztes «crowdsourcing» in den Sozialen Netzwerken des «Web 2.0» gewidmet, damit das Thema des ersten Workshop von 2008 aufgreifend und es weiter entwickelnd. Neben grundsätzlichen Überlegungen und abstrakten Konzepten sind aktuelle Projekte und deren Ansätze für konkrete Applikationen vorgestellt worden, damit dem Generalthema von IRIS 2013, «Abstraktion und Applikation», gerecht werdend.

Entsprechend dem Generalthema von IRIS 2014 – «Transparenz» – ist der thematische Fokus des siebenten Workshop auf die Transparenz von Rechtsetzungsprozessen und ihre Unterstützung durch die neuen Informations- und Kommunikationstechnologien gerichtet gewesen. Die kooperativen Dimensionen von Rechtsetzungsprozessen und ihre Unterstützung durch die neuen Informations- und Kommunikationstechnologien sind im Mittelpunkt des achten Workshop gestanden, das Generalthema von IRIS 2015 – «Kooperation» – aufgreifend. Entsprechend dem Generalthema von IRIS 2016 hat sich der neunte Workshop schwerpunktmäßig mit dem Konzept bzw. den Konzepten des «Netzwerks» befasst, dabei Computernetzwerke, wie sie zur technischen Unterstützung des Rechtsetzungsprozesses erforderlich sind, ebenso berücksichtigend wie die sozialen Netzwerke, in welche Rechtsetzungsprozesse eingebettet sind, und schließlich das Rechtssystem selbst mit den Methoden der Netzwerkanalyse betrachtend. Der zehnte Workshop hat, ganz dem Generalthema von IRIS 2017 entsprechend, das im Blick auf die mittlerweile zwanzigjährige Tradition der Konferenzserie «20 Jahre: Trends und Communities der Rechtsinformatik» gelautet hat, Zwischenbilanz über die Entwicklung der elektronischen Rechtsetzung und des elektronischen Parlaments in Österreich gezogen und Themen aufgegriffen, die im Laufe des vorangegangenen Jahrzehnts bereits aus verschiedenen Perspektiven behandelt worden sind. Entsprechend dem zweigeteilten Generalthema von IRIS 2018, «‹Legal Tech›/Datenschutz», hat sich schließlich der elfte Workshop insbesondere informationstechnischen Anwendungen, die der Unterstützung der Rechtsetzung dienen, gewidmet, und zwar sowohl mit Perspektive auf ihre Gegenstände als auch auf ihr Verfahren. Die zwölfte Ausgabe im Rahmen von IRIS 2019 ist zum einen im Zeichen aktueller Ansätze gestanden, die unter der Bezeichnung «Legal Analytics» zusammengefasst werden können, und hat zum anderen in seminarartiger Form Gelegenheit geboten, Akoma Ntoso bzw. LegalDocML als Standard für die Erzeugung und Verarbeitung legislativer und normativer Dokumente kennenzulernen.

Daran knüpft der dreizehnte Schwerpunkt unmittelbar an, indem er aktuelle Entwicklungen der Rechtsetzungstechnologie vorstellt, denen dieser Standard zugrunde liegt. Zuvor freilich befasst sich der Schwerpunkt in Anknüpfung an das Generalthema von IRIS 2020 mit dem verantwortungsvollen Einsatz von Künstlicher Intelligenz (im weitesten Sinn) in der Rechtsetzung. Neben grundsätzlicher Erörterung möglicher Anwendungsfelder wird vor allem auf Aspekte des europäischen Projekts «ManyLaws» eingegangen, die für die Rechtsetzung relevant sind, insbesondere den Einsatz der in diesem Projekt zu entwickelnden Dienste in der legislativen Entscheidungsunterstützung.

Artificial Intelligence is both a challenge to law-making and a potential means for law-makers. The «Responsible AI» movement, over the past years, has been raising awareness of the potential threats of AI to a «healthy development» of society, and there is an ongoing discussion to what extent normative regulations are required to control the use of AI. The California Legislature, e.g., in 2018, has formally adopted the Asilomar Principles on beneficial AI, thus laying the foundation for a regulatory framework taking into account values like information privacy, accountability, judicial transparency, and respect for human dignity. At the level of the European Union, even though the European Parliament has undertaken some initiative towards preparing a civil-law framework for robotics, the focus so far has been more on discussing ethical than regulatory approaches. The High Level Group on AI, in April 2019, has published ethics guidelines for trustworthy AI. At the same time, in a highly competitive global business environment, there is a strong demand for politically supporting AI development rather than legally restricting it.

The law-making process, in an informed society, is facing the requirement to be «evidence-based», factual evidence being deemed the prerequisite to guarantee the quality of legislation. On the other hand, the elementary concept of democracy to have the political will of the people represented in the law-making process, implies a potential tension between irrational political conviction and rational, evidence-based argument. In this context, the potential of AI algorithms to influence or determine – or «manipulate» – information flows and information selection affects both the irrational and the rational foundations of law-making. The more data is available – and the amount of data produced is growing exponentially –, the more the support of machines is required to turn data into information, and finally into knowledge. The responsible use of these AI based services requires digital literacy, whose promotion therefore is one of the major societal and political challenges. For law-makers, specific tools will be required that, in a transparent manner, would allow them, in particular, to assess the potential impact of policies and normative regulations under consideration, as well as evaluating the actual impact of regulations adopted, with regard to the normative impact, i.e. the impact on the normative system, as well as the societal impact, i.e. the impact on the socio-economic system.

Standardization is a typical and important step in the technology life cycle, affecting both innovation and diffusion, and generally indicating an advanced stage of technology development. In the domain of «Legislative Tech», standardization meanwhile also has taken place, not only in the sense of informal standards emerging but even in the sense of the formal promulgation of a standard by a standardizing body, namely the issuing of Akoma Ntoso or LegalDocML as a standard by OASIS in 2017. This standard defines a set of technology-neutral electronic representations of legislative, legal, and judicial documents, and it includes an XML schema, a naming convention, and a metadata convention.

Metadata being a key resource to semantic technologies, when it comes to improving the access to legal and legislative documents manually edited as well as machine generated metadata will play a key role, which is why metadata conventions and metadata standards demand attention. The «ManyLaws» project, which is funded by the Connecting Europe Facility, intends to build a portal that, within the framework of the European Data Portal (EDP), is to provide transjurisdictional and translingual access to legal and legislative documents, whose metadata are going to be harvested as well as enriched by AI based algorithms applied on the documents within an HPC environment. Thus, the «ManyLaws» metadata convention needs to be based on the relevant metadata standards; actually, three standard metadata schemas will have to be mapped:

• The DCAT Application profile for data portals in Europe (DCAT-AP) is a specification based on W3C’s Data Catalogue vocabulary (DCAT) for describing metadata of public sector datasets in Europe. It is the standard used by the EDP.
• The European Legislation Identifier (ELI) is based on a voluntary agreement between the EU countries. It includes technical specifications on web identifiers (URIs) for legal information, metadata specifying how to describe legal information, and a specific language for exchanging legislation in machine-readable formats.
• Akoma Ntoso has been standardized by the LegalDocML Technical Committee within OASIS. While the ELI metadata schema is based on a comparative analysis of the legal documents of the jurisdictions of the EU and its member states, Akoma Ntoso has its basis in a world-scale comparison of legislative, legal, and judicial documents; thus, it is more comprehensive than ELI, and has a particular focus on the life cycle of legal documents.

Given the variety as well as the relevance of legislative and legal documents, mapping the standards will create an improved framework for describing them and thus be of importance even beyond the scope of the «ManyLaws» project. Thus, it deserves a clear and transparent methodological approach.

Decisions are always made under conditions of uncertainty. Basing them on evidence means reducing the level of uncertainty. Law-makers are facing three dimensions of uncertainty: first, uncertainty about the reaction of the legal subjects, and in democracy of the voters, on legislation to be adopted; second, about the impact of a piece of legislation on the legal system, in particular about possible conflicts with existing regulations, which legally would be settled by derogation, in case of affected regulations of the same or of a lower level in the hierarchy of norms; third, about the impact of a regulation on the socio-economic system, including its diverse sub-systems, which on the one hand would mean whether a regulation was adequate to achieve its political objective, and on the other hand whether it would have unintentional side effects.

Improving the working environment of law makers is one of the aims of the European project «Many-Laws», which is particularly due to the circumstance of two Parliaments, i.e. the Hellenic and the Austrian one, being part of the project consortium. The main objective of the project being to establish a portal allowing for AI supported transjurisdictional and translingual access to legislation at the level of the EU and its member states, the main focus of the decision support system for law-makers will be on enabling them to compare draft regulations with existing regulations, and to assess potential conflicts between them. Thus, even when only focusing on the respective national jurisdiction, law-makers will be able to better assess the potential impact of draft regulations on the legal system, and identify existing norms affected by an intended regulation. Another usage scenario is the transposition of EU legal acts into national legislation; the system will make it easier to refer to already existing or prepared transposition acts in other member states, but it will also analyse the cross-references between the (draft) transposition act(s) and the respective EU legal act to be transposed.

Besides the focus on normative impact assessment, the «ManyLaws» decision support system, according to the project concept, will additionally also offer functionalities to evaluate the public discourse, at least as far as taking place in online social media, on legislative issues under deliberation, so to allow political decision-makers to better take into account the positions taken in civil society on the respective legislative matters.

Over the last decade the legal informatics community has spent huge effort to define technical standards for modelling legal documents (Akoma Ntoso, LegalRuleML, ELI, ECLI, etc.) with the aim to support open access, reusability, interoperability, searchability of legal knowledge in the Internet. Several official gazettes dematerialized their publications (AKN4EU, AKN4UN, Luxembourg, The National Archives UK, Library of Congress of Chile, The US Code, GPO US, etc.), and Parliaments (Senate of Italy, European Parliament, European Commission, UN, etc.) started to use XML techniques for annotating legislative sources to enhance the internal workflow and to improve the law-making process from the beginning.

In this scenario the courts and tribunals, and in general the judiciary system, are one step behind in the standardization of the legal knowledge and documentation. For this reason, this paper presents the conversion of all the judgements of the Constitutional Court of Italy in Akoma Ntoso (http://bach.cirsfid.unibo.it/ldms-cortecostituzionale/) reusing the Open Data official portal material, integrating ECLI and metadata in the unique XML document and implementing usability specifications emitted by the Italian government for the public administration portals.

Additionally, the same UNIBO-CIRSFID research group, with the help of the scholars of European Tax Law, realized a portal of abstract (summary) reports of the European Court of Justice judgments enriched (http://bach.cirsfid.unibo.it/node/massimeTributarie/) with ELI and ECLI metadata extracted using the Open Data service CELLAR of the European Publication Office (https://data.europa.eu/euodp/en/data/group/api).

These two applications are good examples how it is possible to reuse ELI and ECLI URI and metadata, and how to map them in a unique international standard like Akoma Ntoso that is capable i) to manage both legislative and judiciary documents in a unique XML format; ii) to include different ontological and naming convention levels; iii) to create a distributed database among different collection of legal documents; iv) to implement navigation of the references; v) to design advanced searching, visualization and legal data analytics applications.

In 2010 the European parliament launched the eParliament corporate programme to adopt a structured and semantic-driven approach to drafting texts. This marked a clear break from the long standing template and layout based paradigm.

The wide scope of the programme requires active participation of various functional stakeholders within the European parliament. The programme is governed by a steering board composed of high profile officials from the directorate generals of the Presidency, Internal and External Policies, Translation, and Innovation and Technology.

From a technical and architectural standpoint, the implementation is based on a LegalDocML XML dialect and on a limited number of enabling components: a central content repository, a PDF and Word rendering engine, digital signature services solution and an integration services layer.

Finally, a growing suite of business components is being developed on top of the enabling architecture. These components allow the drafting of amendments and parliamentary reports, legal and linguistic verification of texts and translation in any of the 24 EU official languages.

The latest addition to the eParliament business component suite is the Trilogue Table Editor, a highly specialised solution supporting the trilogue negotiations process which is at the core of the Ordinary Legislative Procedure, introduced by the Lisbon Treaty.

The first part of this presentation introduces the eParliament programme, showcasing some of its successes, challenges, setbacks and lessons learned. The second part will focus on the inter-institutional aspects of the programme and specifically on the Trilogue Table Editor project.

In 2019 the Federal Ministry of the Interior commissioned a customization of Akoma Ntoso (titled LegalDocML.de) as a common data format for legislative drafting. Due to complex requirements and the need for stakeholder feedback a model-driven development approach was chosen, in which a formal data model is used to capture the structural requirements for legislative documents, and to validate these with subject matter experts. Technical artefacts and documentation are then generated algorithmically from that model. This talk presents the project context, benefits of model-driven development, and lessons learned from the project.