IA4FriLex: Enhancing The Legislative Consultation Process With AI

Ornella Vaccarelli, Beat Wolf, Talia B. Kimber, Thomas Christin, Thomas Kadelbach
Proceedings of the Fourth Swiss AI Days, PMLR 309:1-12, 2026.

Abstract

Legislative consultation procedures are a core component of participatory law-making but require public administrations to process large volumes of heterogeneous and predominantly unstructured submissions under strict procedural constraints. This task is particularly demanding in practice and often constitutes a bottleneck in legislative workflows. To overcome this challenge, we present IA4FriLex, an AI-assisted pipeline designed to support the processing and synthesis of consultation submissions through a structured and legally grounded workflow. Built exclusively on open-source software and Large Language Models (LLMs), the system automates well-defined stages of consultation handling while preserving human oversight and legal responsibility. IA4FriLex produces standardized, department-ready consultation reports aligned with established cantonal administrative practices. The system has been implemented and deployed in collaboration with the Cantonal administration of Fribourg and evaluated on four real cantonal legislative cases, covering both completed and ongoing consultations. Results show that IA4FriLex reliably generates high-quality first-pass syntheses and reduces consultation report preparation time by at least 80% compared to fully manual drafting. These findings demonstrate that carefully constrained, on-premise deployments of LLM-based systems can effectively support legislative consultation processes, offering a scalable and institutionally compatible approach to AI-assisted law-making.

Cite this Paper

BibTeX
@InProceedings{pmlr-v309-vaccarelli26a, title = {IA4FriLex: Enhancing The Legislative Consultation Process With AI}, author = {Vaccarelli, Ornella and Wolf, Beat and Kimber, Talia B. and Christin, Thomas and Kadelbach, Thomas}, booktitle = {Proceedings of the Fourth Swiss AI Days}, pages = {1--12}, year = {2026}, editor = {Kucharavy, Andrei and Delgado, Pamela and Schürch Todeschini, Valérie and Rumley, Sébastien}, volume = {309}, series = {Proceedings of Machine Learning Research}, month = {23--25 Mar}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v309/main/assets/vaccarelli26a/vaccarelli26a.pdf}, url = {https://proceedings.mlr.press/v309/vaccarelli26a.html}, abstract = {Legislative consultation procedures are a core component of participatory law-making but require public administrations to process large volumes of heterogeneous and predominantly unstructured submissions under strict procedural constraints. This task is particularly demanding in practice and often constitutes a bottleneck in legislative workflows. To overcome this challenge, we present IA4FriLex, an AI-assisted pipeline designed to support the processing and synthesis of consultation submissions through a structured and legally grounded workflow. Built exclusively on open-source software and Large Language Models (LLMs), the system automates well-defined stages of consultation handling while preserving human oversight and legal responsibility. IA4FriLex produces standardized, department-ready consultation reports aligned with established cantonal administrative practices. The system has been implemented and deployed in collaboration with the Cantonal administration of Fribourg and evaluated on four real cantonal legislative cases, covering both completed and ongoing consultations. Results show that IA4FriLex reliably generates high-quality first-pass syntheses and reduces consultation report preparation time by at least 80% compared to fully manual drafting. These findings demonstrate that carefully constrained, on-premise deployments of LLM-based systems can effectively support legislative consultation processes, offering a scalable and institutionally compatible approach to AI-assisted law-making.} }
Endnote
%0 Conference Paper %T IA4FriLex: Enhancing The Legislative Consultation Process With AI %A Ornella Vaccarelli %A Beat Wolf %A Talia B. Kimber %A Thomas Christin %A Thomas Kadelbach %B Proceedings of the Fourth Swiss AI Days %C Proceedings of Machine Learning Research %D 2026 %E Andrei Kucharavy %E Pamela Delgado %E Valérie Schürch Todeschini %E Sébastien Rumley %F pmlr-v309-vaccarelli26a %I PMLR %P 1--12 %U https://proceedings.mlr.press/v309/vaccarelli26a.html %V 309 %X Legislative consultation procedures are a core component of participatory law-making but require public administrations to process large volumes of heterogeneous and predominantly unstructured submissions under strict procedural constraints. This task is particularly demanding in practice and often constitutes a bottleneck in legislative workflows. To overcome this challenge, we present IA4FriLex, an AI-assisted pipeline designed to support the processing and synthesis of consultation submissions through a structured and legally grounded workflow. Built exclusively on open-source software and Large Language Models (LLMs), the system automates well-defined stages of consultation handling while preserving human oversight and legal responsibility. IA4FriLex produces standardized, department-ready consultation reports aligned with established cantonal administrative practices. The system has been implemented and deployed in collaboration with the Cantonal administration of Fribourg and evaluated on four real cantonal legislative cases, covering both completed and ongoing consultations. Results show that IA4FriLex reliably generates high-quality first-pass syntheses and reduces consultation report preparation time by at least 80% compared to fully manual drafting. These findings demonstrate that carefully constrained, on-premise deployments of LLM-based systems can effectively support legislative consultation processes, offering a scalable and institutionally compatible approach to AI-assisted law-making.
APA
Vaccarelli, O., Wolf, B., Kimber, T.B., Christin, T. & Kadelbach, T.. (2026). IA4FriLex: Enhancing The Legislative Consultation Process With AI. Proceedings of the Fourth Swiss AI Days, in Proceedings of Machine Learning Research 309:1-12 Available from https://proceedings.mlr.press/v309/vaccarelli26a.html.

Related Material