ALPINE

ALPINE

AI Litigation and Procedural Fairness: A Needs-Based Approach to Evidence Access

Ljupcho GROZDANOVSKI — Julien CABAY — Ashwin ITTOO

With the support of the Action de Recherche Concertée (ARC) program at ULiège and centered on procedural fairness in AI-related disputes, the ALPINE project (48 months) aims to identify the procedural mechanisms that should be made available to litigants in cases involving AI technologies, in order to enhance their effective access to legal processes and remedies.

Following the guiding thread of effective or meaningful litigant participation, ALPINE pursues three key objectives: first, it will empirically identify the features and procedural conditions of effective litigant participation  in AI-related disputes. Second, and based on the findings of the empirical analysis, it will theoretically develop a new subfield within fairness scholarship, proposing a needs-based model for procedural fairness in cases involving AI-related harm and causation. Finally, it will practically implement these findings through the creation of an online tool designed to help stakeholders anticipate and address key procedural and adjudicatory challenges related to the AI systems they develop or deploy.

These objectives will be pursued through three interconnected Work Packages (WPs). WP1 will translate to a public consultation across five EU Member States (Belgium, France, Italy, Germany, and the Netherlands), as well as with institutions at the EU and Council of Europe/ECHR levels. Expert respondents - incl. lawyers, judges, academics, and data scientists - will be asked to identify the types of evidence necessary to substantiate claims about harm and causation in AI disputes. The data collected from the consultation will allow the ALPINE team to map out the key categories of evidence (expert reports, testimonies, data disclosure, etc.) stakeholders perceive as essential and most probative, for the effective participation and judicial redress in AI-related litigation. WP2 builds on the findings of WP1 by assessing whether the legal frameworks in the selected jurisdictions support access to the types of evidence identified. Through a black-letter analysis of national procedural laws, relevant EU law provisions, the ECHR, and corresponding case law, the research carried out in this WP will clarify the procedural safeguards - such as access to evidence, equality of arms, adversarial debate, and trade secret limitations - needed to practically ensure the right to a fair trial in current and emerging AI liability cases.

WP3 will translate the insights from WPs 1 and 2 into an accessible ALPINE Procedural Compliance Checkpoint - an online toolkit aimed to equip stakeholders with practical guidance on procedural norms, relevant case law, derogatory regimes (regarding evidence disclosure), and intersecting rights such as IP and trade secrets, thereby bridging research findings and real-world applications.

As a result of its three-phase interdisciplinary approach, ALPINE’s main doctrinal and societal contribution lies in the (re)conceptualizing of "effective" and "meaningful" participation in the context of AI litigation, feeding directly into recent and flourishing regulatory and scholarly debates on procedural fairness in AI governance.