AI in the Service of Justice: Opportunities and Threats in the Application of Language Models (LM) and Large Multimodal Models (LMM) in Judicial and Extrajudicial Dispute Resolution with Specific Consideration of Civil Law. Reflections on the Current
AI in the Service of Justice: Opportunities and Threats in the Application of Language Models (LM) and Large Multimodal Models (LMM) in Judicial and Extrajudicial Dispute Resolution with Specific Consideration of Civil Law. Reflections on the Current

Summary/Abstract: The Polish justice system is hampered by the protracted nature of court proceedings, urgently requiring a reassessment of its current model. The paper examines the potential benefits and severe risks of integrating AI and LLMs into dispute resolution. The core opportunity lies in data-driven optimization, offering greater efficiency, standardization of jurisprudence, and administrative support (e.g., the digital judge’s assistant). However, the implementation of AI as an adjudicator faces critical threats: ethical issues surrounding the dehumanization of law and the inability to interpret human-centric concepts such as "good faith"; technical risks from "black box" algorithms and the replication of historical biases; and sovereignty concerns tied to strategic infrastructure and compliance with the EU AI Act. While AI is not ready to replace human judges, its inevitable role in supportive functions and ADR mandates immediate structural reform to leverage its benefits while safeguarding the constitutional principles of impartiality, human dignity, and the right to a fair trial.

• Details
• Contents