Project SURAKSHA: A Zero-Trust, Leak-Resilient and Publicly Verifiable Examination Architecture for NEET

Investigative reporting has documented a persistent examination-security problem in India: one Indian Express investigation counted 41 recruitment-exam leaks across 15 states over five years, affecting about 1.4 crore applicants, while an India Today compilation counted more than 70 leak cases across seven years affecting about 1.7 crore aspirants; the two figures use different scopes and should be read as orders of magnitude rather than an official census. NEET-UG provides two instructive failures. In 2024, the Supreme Court held that a leak had occurred at Hazaribagh and Patna but found insufficient material to conclude that the breach was systemic enough to cancel the national examination. In 2026, the CBI registered a fresh case concerning unauthorized pre-exam circulation; it subsequently alleged that NTA-associated subject experts had disclosed live Chemistry and Biology questions, and NTA conducted a nationwide re-examination on 21 June 2026. Project SURAKSHA responds to the structural lesson: no single actor should have routine pre-session access to a complete, high-value national paper. The proposed seven-layer architecture combines (1) a large calibrated item bank and constrained computer-based form assembly, (2) a secure on-site-printing bridge where paper remains necessary, (3) two-person integrity and separation of duties, (4) forensic traceability, (5) exam-day controls with privacy and due-process safeguards, (6) durable legal and institutional governance, and (7) public verifiability through cryptographic commitments and auditable logs. A dedicated psychometric Fairness Gate addresses the central objection to different candidate forms: forms may contain different questions, but they must satisfy the same content blueprint and predeclared psychometric tolerances, with independent audit and staged rollout. The paper also presents an illustrative leak-economics sensitivity model; its probabilities and monetary consequence-equivalents are scenario assumptions, not forecasts. SURAKSHA is a proposal for independent evaluation and pilot testing, not a claim of zero leak probability or a ready procurement specification.