examining the potential challenges | 10.1163/22131035-14020001 4 27 Reliability of ai 4.1 Navigating the Complexities of Nationality Laws Another challenge is that determining a person is stateless involves its own peculiarity and complexity that might not be grasped by ai algorithms. Statutory laws and international treaties often require human interpretation where there are ambiguous and normative or context-dependent meanings.154 ai systems may lack the ability to interpret complex legal terms that human adjudicators provide and may confuse the meaning of certain terms because the technological capacity to readily code complex legal text still does not exist.155 The European Union (EU) Artificial Intelligence Act has labelled as ‘high risk’ ai systems intended to be used by a judicial authority in interpreting facts and the law and applying the law to facts or by alternative dispute resolution when the outcomes produce legal effects for the parties.156 In certain situations, the intricacy of a state’s nationality laws and practices may warrant the use of human expert evidence.157 Consequently, the depth of understanding that humans can provide remains indispensable. Olivier argues that ‘human intelligence in the decision-making process transcends logic and is not entirely based on some sort of specific algorithm process’.158 Humans possess the ability to engage in critical thinking and nuanced reasoning – capacities that ai currently lacks. Given that ai is not yet sufficiently advanced to interpret complex legal terminology, it should not replace human decision-makers. I now elaborate upon some of the complexities involved in the statelessness area. In state’s nationality legislation, the term ‘national’ is used differently in different countries and different nationals could be granted distinct rights.159 If ai has to determine statelessness, how ai algorithms will navigate the different terms used in different countries is difficult to understand. ai may not perform well, or at all, in areas that involve abstract concepts such as ‘reasonableness’.160 This concept involve actually understanding the 154 155 156 157 158 159 160 Magnus Killander, ‘Interpreting Regional Human Rights Treaties’ (2010) 7 (13) sur International Journal on Human Rights 145–169. Foster, (n 6) at 6. Veronika Fikfak and Laurence R Helfer, ‘Automating International Human Rights Adjudication’ (2025) 69 46 Michigan Journal of International Law 117; European Union Artificial Intelligence Act 2024 Annex iii, point 8. Handbook on Statelessness (n 13) para 85. Olivier, (n 133). Handbook on Statelessness,(n 13) para 52: “citizen”, “subject”, “national” in French, and “nacional” in Spanish. Harry Surden, ‘Artificial intelligence and law: An overview’ 35(4) 2019 Georgia State University Law Review 1326. International Human Rights Law Review (2025) 1–31