Ethics of Using Artificial Intelligence in the Analysis of Possession for Usucapion

Abstract

This paper analyzes the complexity and ethical implications of using artificial intelligence (AI) in the legal process of establishing possession, as a fundamental element of usucapion. In Romanian civil law, usucapion represents an original way of acquiring property, conditioned by a useful, continuous, uninterrupted, undisturbed possession and under the name of the owner, according to art. 930 of the Civil Code. The assessment of these conditions is, traditionally, the attribute of the court, but the emergence of AI-based technologies introduces new ways of factual and evidentiary analysis, which require an adequate ethical and legal framework. The use of AI in the field of law, including in the context of usucapion, promises to increase efficiency by analyzing large volumes of data – such as satellite images, cadastral archives or video recordings – and by corroborating information obtained from various sources. However, these advantages are counterbalanced by significant ethical risks: algorithmic errors, lack of decision-making transparency and the possibility of systemic discrimination caused by incomplete or biased data. A central aspect of the ethical approach is the risk that algorithms may misinterpret the nature of possession – for example, considering sporadic presence as continuous possession. Such errors can seriously affect the recognition or rejection of a property right. Furthermore, the use of AI in legal proceedings calls into question the respect of the fundamental principles of the right to defense and adversarial proceedings, especially when the algorithms cannot be audited or explained (so-called “black box algorithms”). Another crucial element is the compatibility of the use of AI with the right to privacy. Since the analysis of possession frequently involves the use of personal data – images, digital footprints or information from social networks – it is essential that this practice complies with the legislation on the protection of personal data (GDPR), and the case law of the ECHR. In legal terms, evidence generated or interpreted by AI must comply with the conditions of admissibility and loyalty of evidence, as provided for in the Code of Civil Procedure. Any evidence administered must be able to be contested, interpreted and discussed equally by the parties, which is not always guaranteed in the case of opaque or insufficiently documented AI. The paper therefore proposes a series of recommendations: using AI only as an assistive tool, not a decision-making one; auditing and external validation of the algorithms used; establishing a clear right to contest the results generated by AI; and developing a specific regulatory framework to regulate the use of automated technologies in the field of real rights. In conclusion, although AI can bring important benefits in the management and assessment of possession for usucapion, these advantages must be balanced by rigorous legal and ethical control, in order to protect the fundamental rights of the person and to maintain the integrity of the act of justice.