The integration of artificial intelligence (AI) into the legal domain is transforming traditional legal practices in the legislative domain by enhancing efficiency and accuracy. Nevertheless, the deployment of AI within Parliaments, Assemblies, and Governments necessitates careful consideration of the constraints imposed by the theory of law, as well as adherence to democratic principles and ethical values. These dimensions should be incorporated by design into the entire lifecycle of AI system development and maintenance, through mechanisms such as human-in-the-loop, human-on-the-loop, and human-in-command approaches. Establishing a robust conceptual and regulatory framework in this area is essential to guide technical experts, private AI providers collaborating with deliberative institutions, and the bodies responsible for digital transformation. Scholars in legal philosophy, digital constitutionalism, AI law, and digital ethics therefore play a crucial role in guiding the evolution of legal theory to ensure that it adapts coherently to technological change while preserving core democratic principles. A critical challenge lies in preserving the autonomy of decision makers and the separation of powers amid digital transformations, while ensuring that silence, dissent, and the plurality of legal interpretation are meaningfully represented and not constrained by the rigid codification of norms within AI systems. Understanding these advancements is essential not only for legal professionals, who must remain both legally and ethically focused, but also for computer scientists and engineers, who are responsible for designing and implementing solutions that effectively support institutional functions. This Dagstuhl Seminar aims to explore this interdisciplinary domain using five main topics:

1. AI and Digital Legislation Framework: What aspects of the legislative process can be reasona-bly subject to automation, and what potential can be harnessed through the use of hybrid AI ap-proaches (symbolic, sub-symbolic, neuro-symbolic)? What methodological frameworks are re-quired to conduct research in this domain without risking the fundamental pillars of democracy?
2. Legal Language in the LLM Era: How to modify the AI models to the future society that we want, and not only based on a stochastic approach? How to also inject the semantic web tech-nologies and the local parameters coming to the legal tradition and the domestic legal culture?
3. Theory of Law and Ethical Issues in the AI Era: Which new legal theory is necessary to cope with the challenge of AI in the deliberative institutions? How is the power moving with the use of AI in the law-making process?
4. Democracy and Deliberative Institutions in the AI Era: How can democratic and deliberative institutions preserve human autonomy and legitimacy while integrating AI into legislative and policy-making processes? How can AI systems be designed to ethically and accurately represent silence, dissent, and other nuanced elements of democratic discourse, and what safeguards are necessary to prevent bias, manipulation, and lack of transparency?
5. Official Statistics Institutions and Data Analytics to Measure the Law Effect: How can sta-tistical institutions and data analytics be used to evaluate the impact of laws on society by identi-fying patterns, measuring outcomes, and supporting evidence-based policymaking while improv-ing transparency and detecting emerging social needs?

This Dagstuhl Seminar aims to foster interdisciplinary dialogue by bringing together diverse scholarly communities to develop a shared and robust framework for the use of AI in deliberative institutions.

Creative Commons BY 4.0

Monica Palmirani, Victor Rodriguez Doncel, Jaromír Šavelka, and Aurelia Tamò-Larrieux