The current situation in relations between Poland and Ukraine, the war events in Ukraine, and the economic cooperation between the two countries have their own legal implications. There is no doubt that contracts concluded between Polish and Ukrainian entities will generate legal proceedings in connection with non-performance or improper performance of contractual obligations by one or both counterparties. Consequently, the laws of both countries governing the recognition and enforcement of foreign judgments will be relevant. It should be emphasized that so far the number of such proceedings in Polish-Ukrainian relations has been negligible, as evidenced by the lack of a large number of court cases in this respect. However, bearing in mind the development of economic cooperation and the need to establish uniform and legally justified judicial practice in the future regarding the recognition and enforcement of Ukrainian court judgments in Poland, judges should be aware of which provisions should be applied in such proceedings and to what extent. The correlation between the provisions of the 1993 Polish-Ukrainian Convention and the domestic law of Ukraine and Poland is not straightforward. The purpose of the present gloss is to present the most controversial aspects related to the application of governing law in commercial cases concerning the recognition and enforcement of foreign judgments, on the example of one of the few judgments issued in this area. In particular, the gloss concerns the analysis of the Polish court’s failure to apply the provisions of the agreement between Poland and Ukraine on legal assistance and legal relations in civil and criminal matters dated 24 May 1993, and the legal consequences of applying domestic law instead.

This paper examines the functioning of large language models (LLMs) in the field of private international law. Its primary objective is to assess their effectiveness in determining the applicable law, particularly in the analysis of choice-of-law clauses, and to identify structural limitations arising from the architecture of these models. The study applies a dogmatic-legal and functional methodology, drawing on examples from legal practice, including due diligence processes and the use of AI systems in law firms. The central thesis is that the challenges associated with LLMs in private international law are structural in nature and stem from their probabilistic design, including tokenization, hallucinations, and semantic averaging. These features create a risk of misidentifying conflict-of-law rules, misinterpreting choice-of-law clauses, and introducing systemic bias toward dominant legal systems. The conclusions indicate that despite significant technological progress, particularly in reasoning models and retrieval-augmented generation (RAG), LLMs cannot replace legal professionals in applying conflict-of-law rules. Their use requires continuous human oversight, and the development of specialized, locally trained models may provide a more suitable path for private international law practice.