How Two Independent Researchers Are Building India's AI Safety, Governance and Cybersecurity Stack

Counts differ because each database indexes a different set of publications. We treat OpenAlex as the canonical count; Google Scholar is not shown (no API, and crawling it violates its ToS).

This technical note explains, at overview level, a portfolio of Indian patent applications (filed, pending, not granted) for a layered AI Safety, Governance and Cybersecurity stack. Its common thread is signed, tamper-evident and independently verifiable evidence across six layers: hardware silicon roots of trust, physical systems, a certified AI foundation (AI Core), runtime defence (Universal Immune System), behavioural safety (Deep Intent Mapping Framework), and compliance and identity (Overlay Framework and biometric de-duplication governance). The article includes a hardware-enforced financial compliance engine. It maps the design to India's policy landscape, including the India AI Governance Guidelines (MeitY), the Information Technology Act, 2000 and CERT-In directions, the Digital Personal Data Protection Act, 2023, RBI digital payment standards and Income Tax Department reporting requirements, and to constitutional principles developed by the Supreme Court of India. It states plainly what is buildable now, what is research-stage, and what remains design intent, and it proposes independent validation through national shared facilities such as I-STEM. Named institutions are referenced for orientation only; none has sponsored, reviewed or endorsed this work. Nothing here is legal, financial or investment advice.