AI Hallucinations And Safety Framework In India: A Comparative Analysis Of The EU AI Act, NIST AI RMF, And The AI (Ethics And Accountability) Bill, 2025
- IJLLR Journal
- 7 minutes ago
- 1 min read
Debarshi Roy Choudhury, B.A. LL.B., Jalpaiguri Law College
ABSTRACT
India is developing one of the world's largest AI ecosystems, yet it possesses one of the least developed AI-specific legal regimes and this paradox is not merely ironic but constitutionally perilous. The core issue addressed in this article is the absence of binding legal obligations in Indian law concerning algorithmic bias, AI hallucinations, and output reliability harms that are already being experienced on a large scale in critical domains such as credit assessment, judicial risk evaluation, and healthcare diagnostics. Three comparative frameworks are analyzed: the EU AI Act 2024, selected for its risk-based regulatory structure and its implementation of the proportionality principle in AI governance; the NIST AI Risk Management Framework for its technical operationalizability and detailed methodologies for measuring bias and hallucination; and the proposed AI (Ethics and Accountability) Bill, 2025, selected as the most pertinent domestic legislative pathway for examination. The primary argument posited is that India’s current voluntary, principle-based approach to AI governance, as outlined in the MeitY AI Governance Guidelines, 2025, is both constitutionally and practically insufficient when compared to the significant risks posed by AI bias and hallucinations to fundamental rights enshrined in Articles 14 and 21 of the Constitution. This article contributes uniquely by proposing a four-pillar legislative framework: a statutory risk-classification system, mandatory pre- deployment bias audits and hallucination benchmarking, a right to explanation and algorithmic redress, and a technically mandated AI Safety Institute which are all tailored to India’s distinctive constitutional, social, and technological context.
Keywords: Algorithmic Bias, AI Hallucinations, EU AI Act, NIST AI RMF, AI Ethics and Governance, AI Safety, Fundamental Rights
