Can an Algorithm Be a Lawful Manager? Delegated Judgment, Non-Delegable Responsibility in Indian Employment Law

When an employer lets an algorithmic system score, rank or flag workers for decisions that affect their livelihoods, existing legal categories struggle to say who must answer for the result. This article argues that Indian law already contains the resources needed to answer that question, even though no statute or judgment has yet confronted it directly: an employer may lawfully delegate the exercise of managerial judgment to a computational system, but it cannot delegate away legal responsibility for what that system decides. Examining constitutional non-arbitrariness doctrine, the reasoned-decision requirement in Indian administrative law, the newly consolidated Industrial Relations Code, 2020, and the still largely dormant Digital Personal Data Protection Act, 2023, the article shows that India’s existing protections are genuine but fragmented, and that the fragmentation tracks a specific fault line between State and private employment rather than a general absence of principle. Drawing a narrow, non-imitative lesson from the European Union’s own unsettled AI Act, the article proposes a proportionate, consequence-based framework: meaningful human review, decision-relevant (not source-code) explanation, data accuracy and contestability, and non-delegable employer accountability, built substantially from institutions Indian labour law already possesses.