Accountability and Transparency in AI Systems: A Human Rights Perspective

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

The proliferation of artificial intelligence across public and private sectors has positioned algorithmic decision-making at the heart of consequential outcomes, from criminal sentencing and credit scoring to healthcare allocation and employment screening. Yet the opacity of AI systems, combined with fragmented governance structures, has created a significant accountability deficit that threatens the realisation of fundamental human rights. This paper addresses the central research question: to what extent do existing human rights legal frameworks impose enforceable accountability and transparency obligations on states and corporate actors deploying AI systems? Using a doctrinal legal analysis methodology, the paper systematically examines primary legal instruments, including the UN Guiding Principles on Business and Human Rights, the International Covenant on Civil and Political Rights, the EU AI Act, and UNESCO’s Recommendation on the Ethics of AI, to assess their capacity to govern AI-driven decision-making in rights-sensitive contexts. With special reference to the Indian constitutional framework, anchored in the Supreme Court’s landmark recognition of privacy as a fundamental right in Justice K.S. Puttaswamy (Retd.) v. Union of India (2017), and to the emerging architecture of the Digital Personal Data Protection Act, 2023, the paper situates India’s governance challenges within the global accountability debate. The paper identifies critical governance gaps, including the absence of enforceable audit mechanisms, inadequate redress pathways for rights-affected individuals, and the difficulty of attributing legal responsibility across complex AI supply chains. It concludes by proposing a human rights-by-design governance model that embeds accountability and transparency obligations across the full AI lifecycle, offering a principled framework for reconciling technological innovation with fundamental rights protection.