Faith Framework (v1.0) - Unified Agentic AI Governance for the Indian Financial Sector

Abstract

As Banks, Non-Banking Financial Companies (NBFCs), and Fintechs in India transition from predictive machine learning to autonomous Agentic Artificial Intelligence (AI), AI governance professionals face an acute operational bottleneck: regulatory fragmentation. Deploying autonomous agents requires reconciling the RBI FREE-AI Framework (2025), the DPDP Act (2023), the NIST AI RMF (1.0), and OWASP LLM controls (2025). This paper introduces The FAITH Framework (v1.0) (Financial AI Trust & Human-Control Framework), a unified operational meta-framework tailored for Indian financial institutions. FAITH v1.0 distills these intersecting regimes into five enforceable pillars (Financial Controls First, Accountable Data Use, Interpretable by Design, Trust but Continuously Verify, and Humans Stay Accountable), complete with a 15-control catalog, 3LoD implementation guidance, and evidentiary audit mappings for RBI supervisory compliance.

Governing Agentic AI in Africa: A Systematic Literature Review of Ethical Challenges, Framework Gaps, and Ubuntu-Informed Governance Solutions
Tope Samuel Adeyelure
Journal of Information Systems and Informatics24 Oct 2026Peer-reviewedCC BYFull textGeneral mathematicsDOAJ
On the Multi-Index, Multi-Rate, and Multi-Phase Dynamics of Decoder-Only Language Models: A Unified Hybrid Framework for Generative and Agentic Systems
Ali Pakniyat
arXiv1 Oct 2026PreprintarXiv licenceFull textOptimisation & controlarXiv
The Point of Creation: A Unified Framework for Generative Self-Governance
A.M. Sterling
Zenodo17 Oct 2026PreprintCC BY 4.0Zenodo
ESG disclosure and financial performance in Indian hospitality firms using text mining
Dileep G. Menon, Sangeetha Gunasekar, Badri Narayanan Gopalakrishnan
Discover Sustainability1 Jul 2026Peer-reviewedCC BYFull text1 citedEnvironmental science & pollutionDOAJ
From Compliance-Based to Collaborative Financial Governance: A Bibliometric-Systematic Review of Multi-Actor Accountability in Public Financial Management.
Simon Saimima, Andy Fefta Wijaya, Endah Setyowati +1
F1000Research2 Nov 2026Peer-reviewedCC BYFull textEurope PMC

Cite

BibTeX RIS

BibTeX

@misc{khan2026faith,
  title = {{Faith Framework (v1.0) - Unified Agentic AI Governance for the Indian Financial Sector}},
  author = {Asif Khan},
  howpublished = {Zenodo},
  year = {2026},
  month = oct,
  doi = {10.5281/zenodo.23538748},
  url = {https://zenodo.org/records/23538748},
  note = {CC BY 4.0},
}

RIS

TY  - UNPB
TI  - Faith Framework (v1.0) - Unified Agentic AI Governance for the Indian Financial Sector
AU  - Asif Khan
PY  - 2026
DA  - 2026/10/21
PB  - Zenodo
DO  - 10.5281/zenodo.23538748
UR  - https://zenodo.org/records/23538748
AB  - As Banks, Non-Banking Financial Companies (NBFCs), and Fintechs in India transition from predictive machine learning to autonomous Agentic Artificial Intelligence (AI), AI governance professionals face an acute operational bottleneck: regulatory fragmentation. Deploying autonomous agents requires reconciling the RBI FREE-AI Framework (2025), the DPDP Act (2023), the NIST AI RMF (1.0), and OWASP LLM controls (2025). This paper introduces The FAITH Framework (v1.0) (Financial AI Trust & Human-Control Framework), a unified operational meta-framework tailored for Indian financial institutions. FAITH v1.0 distills these intersecting regimes into five enforceable pillars (Financial Controls First, Accountable Data Use, Interpretable by Design, Trust but Continuously Verify, and Humans Stay Accountable), complete with a 15-control catalog, 3LoD implementation guidance, and evidentiary audit mappings for RBI supervisory compliance.
ER  -

Abstract

Related

Cite