Source: Indian Journal of Law and Legal Research

Artificial intelligence is rapidly entering legal practice with capabilities to search law, compare documents, identify inconsistencies, and assist drafting, but the piece stresses that adjudication transcends data processing and requires evaluative human judgment. India's Goods and Services Tax (GST) system produces show cause notices, determinations, penalties and appeals that demand assessments of credibility, intent, and appropriate consequences—features a purely pattern-based AI may fail to appreciate. The article uses statutory cues to underline this point: Section 75(4) of the CGST Act mandates an opportunity of hearing when an adverse decision is contemplated or requested, while Section 75(6) requires the proper officer to state relevant facts and the basis for decisions. An illustrative comparison is offered: two taxpayers with identical reporting errors may differ legally if one deliberately suppressed a transaction and another made a clerical mistake supported by records, a distinction hinging on evidence and surrounding circumstances.

The author argues that AI can and should assist by organising documents, identifying missing records, retrieving authorities, and flagging inconsistencies, but assistance must not be equated with adjudication or used to displace the human role at the centre of fair decision-making. The piece also invokes Supreme Court precedents on natural justice—naming Canara Bank v Debasis Das and Dharampal Satyapal Ltd v Deputy Commissioner of Central Excise—to connect fair hearing with reasoned decision-making and to caution against delegating core evaluative duties to automated systems.