Safeguarding Human Rights in the Global AI Ecosystem: Toward a Unified and Binding International Legal Framework

Abstract

Artificial intelligence has moved from a discrete technical domain to a structural feature of
contemporary governance, economic organization, and social life, reshaping how decisions are
made, how information circulates, and how power is exercised both within and across borders.
This shift carries profound consequences for the international human rights order. Existing
responses, ranging from voluntary ethical charters to binding instruments such as the European
Union's Artificial Intelligence Act and the Council of Europe's Framework Convention on
Artificial Intelligence, Human Rights, Democracy and the Rule of Law, remain regionally
concentrated, doctrinally narrow, and largely silent on the structural and transboundary character
of AI's human rights footprint. This article argues that the international community requires a
single, treaty-based instrument, anchored in the United Nations Charter and the International Bill
of Human Rights, capable of reaching the full life cycle of artificial intelligence systems: their
design, their data and material supply chains, their deployment, and their long-term societal
consequences. The article develops an independent legal and policy argument for why the United
Nations Human Rights Council is institutionally best placed to convene this process, what
substantive pillars such a convention must contain, and how a phased pathway, beginning with a
General Assembly resolution, can realistically lead toward a binding global regime. Attention is
given to groups whose protection existing frameworks have left incomplete, including indigenous
peoples, populations of the Global South, and individuals exposed to algorithmic decision-making
in high-stakes settings such as health, justice, migration, and armed conflict