A Human Rights-Based Approach to Responsible AI

Human rights provide a universal framework for AI ethics: Rather than relying on implicit assumptions about values, the paper argues that universal human rights offer an explicit, globally recognized set of principles that can guide responsible AI development across different cultures and contexts.

Current AI ethics research lacks clear value alignment: The paper identifies a critical gap in existing fairness, accountability, and transparency research—these efforts often assume an implicit consensus on values without making those values explicit, which is problematic in a pluralistic world with diverse value systems.

Shift focus from machines to people: The human rights approach reorients responsible AI discussions away from technical concerns about algorithmic bias toward a human-centered question: "who is harmed, what harms do they face, and how can those harms be prevented?"

Human rights framework enables civil society participation: By grounding AI governance in established human rights doctrine, the paper suggests this approach can facilitate meaningful partnership and participation from civil society organizations, rather than leaving decisions solely to technologists and corporations.

Universal human rights bridge cultural differences: The paper positions human rights as "cross-culturally recognized" values, offering a potential common language for addressing AI's global impacts despite differing cultural and political contexts.