AI Governance in India: A Pragmatic Framework for Regulatory Coherence

India does not need a blank-sheet AI law. Within a single policy cycle, it has assembled a dense body of instruments, the India AI Governance Guidelines, the IT (Intermediary Guidelines) Amendment Rules on synthetically generated information, the DPDP Act and its Rules, sectoral action from the RBI, SEBI, IRDAI and ICMR, the CCI's market study, and the DPIIT copyright process. The binding constraint for the next phase is therefore not a shortage of rules but whether these instruments cohere into a single, predictable system or pull firms in contradictory directions. This paper proposes a framework to supply that coherence. It preserves India's stated philosophy principle-based, techno-legal, innovation-first, with an inclusion floor and adds the connective tissue, legal foundations, and institutional design needed to make it work, built on three strengths only India can fully exploit: its linguistic diversity, its digital public infrastructure, and its capacity to enforce policy through verifiable technical means. Deliberately honest about feasibility, it assumes ideal institutions will under-deliver and designs fallbacks accordingly. Three priority actions unblock the rest: give governance coordination real authority, reconcile the data-protection, synthetic-media and training-data conflict in law, and recast the AI Safety Institute as a standards-and-accreditation body.