GLOBAL GOVERNANCE OF ARTIFICIAL INTELLIGENCE: THE ROLE OF THE UNITED NATIONS AND THE LIMITS OF INTERNATIONAL REGULATION

Abstract

The study addresses the challenges that the accelerated development of artificial intelligence poses to International Law, in view of the transnational nature of data flows, the global operations of the companies that develop the technology and the risks it represents for human rights and democratic processes. It is a theoretical study with a qualitative approach, which aims to analyze the role of the United Nations (UN) in building global governance of artificial intelligence and to assess the potential and the limits of soft law and hard law instruments in the international regulation of the subject. The methodology combines the deductive method with a bibliographic review of works on International Law and Constitutional Law and with a documentary analysis of international instruments, such as the report Governing AI for Humanity, the UNESCO Recommendation on the Ethics of Artificial Intelligence, the UN General Assembly resolution on the subject and the encyclical Magnifica Humanitas, complemented by illustrative cases of disinformation produced with artificial intelligence. The results indicate that national regulation is insufficient to address risks that cross borders and that the UN has acted mainly through non-binding instruments, a choice explained by geopolitical divergences among the technological powers and by the speed of innovation. It concludes that the UN has the legitimacy to lead the construction of international standards for artificial intelligence governance and that soft law is a viable starting point, but that the effective protection of human rights will probably require its complementation by legal mechanisms endowed with binding force.