India's rapid deployment of artificial intelligence — including facial recognition at railway stations, airports, and exam halls — is outpacing its legal framework, with critics warning that vulnerable communities bear the heaviest costs. At February's India AI Impact Summit 2026 in New Delhi, Delhi Police deployed over 4,000 AI-enabled cameras and facial recognition systems across the city to monitor attendees and the public, a setup Amnesty International said exemplified how such technologies restrict civic space and enable state surveillance. The risks are already concrete: a government nutrition programme serving some 47 million women and children introduced AI facial recognition in mid-2025, and by year's end nearly half of intended recipients had been denied food because the system failed to identify their faces — a failure linked to poor lighting, low-end devices, and well-documented bias against darker skin tones. India currently lacks a binding AI law; its November 2025 governance guidelines are non-binding, and a proposed AI ethics bill has yet to be enacted, leaving rights groups to warn that the country's "responsible AI" rhetoric amounts to little more than aspiration without enforceable accountability.