Governed AI Agents in Consumer Banking: Data Stewardship and Data Quality Framework for Credit Card and Deposit Analytics

References

[1] Abraham, R., Schneider, J. and vom Brocke, J. (2019) 'Data governance: A conceptual framework, structured review, and research agenda', International Journal of Information Management, 49, pp. 424-438. doi: 10.1016/j.ijinfomgt.2019.07.008.

[2] Aldridge, I. et al. (2026) 'Agentic artificial intelligence in finance: A comprehensive survey'. Available at: https://arxiv.org/abs/2604.21672.

[3] Axelsen, H., Licht, V. and Damsgaard, J. (2025) 'Agentic AI for financial crime compliance'. Available at: https://arxiv.org/abs/2509.13137.

[4] Basel Committee on Banking Supervision (2013) Principles for effective risk data aggregation and risk reporting. Basel: Bank for International Settlements. Available at: https://www.bis.org/publ/bcbs239.pdf.

[5] Biecek, P. et al. (2021) 'Enabling machine learning algorithms for credit scoring: Explainable artificial intelligence methods for understanding complex predictive models'. Available at: https://arxiv.org/abs/2104.06735.

[6] Bücker, M., Szepannek, G., Gosiewska, A. and Biecek, P. (2022) 'Transparency, auditability and explainability of machine learning models in credit scoring', Journal of the Operational Research Society, 73(1), pp. 70-90.

[7] Consumer Financial Protection Bureau (2022) Consumer Financial Protection Circular 2022-03: Adverse action notification requirements in connection with credit decisions based on complex algorithms. Available at: https://www.consumerfinance.gov/compliance/circulars/circular-2022-03-adverse-action-notification-requirements-in-connection-with-credit-decisions-based-on-complex-algorithms/.

[8] Consumer Financial Protection Bureau (2025) The Consumer Credit Card Market: Report to Congress. Washington, DC: CFPB. Available at: https://files.consumerfinance.gov/f/documents/cfpb_consumer-credit-card-market-report_2025.pdf.

[9] Federal Deposit Insurance Corporation (2025) Quarterly Banking Profile. Washington, DC: FDIC. Available at: https://www.fdic.gov/quarterly-banking-profile.

[10] Federal Reserve Board (2026) Consumer Credit - G.19. Washington, DC: Board of Governors of the Federal Reserve System. Available at: https://www.federalreserve.gov/releases/g19/.

[11] Federal Reserve Board, Office of the Comptroller of the Currency and Federal Deposit Insurance Corporation (2026) Revised guidance on model risk management. Washington, DC: Federal Reserve/OCC/FDIC.

[12] FinRegLab (2025) The next wave arrives: Agentic AI in financial services. Washington, DC: FinRegLab. Available at: https://finreglab.org/research/agentic-ai-in-financial-services/.

[13] Homwe, T., Mupa, M.N., Matope, A., Mlambo, N., Chingezi, L. and Chihota, T.A. (2025) 'Interpretable machine learning for audit planning: Improving misstatement and compliance risk detection in financial services', World Journal of Advanced Research and Reviews, 28(2), pp. 925-933. doi: 10.30574/wjarr.2025.28.2.3779.

[14] Khatri, V. and Brown, C.V. (2010) 'Designing data governance', Communications of the ACM, 53(1), pp. 148-152. doi: 10.1145/1629175.1629210.

[15] Kurshan, E., Shen, H. and Chen, J. (2020) 'Towards self-regulating AI: Challenges and opportunities of AI model governance in financial services'. Available at: https://arxiv.org/abs/2010.04827.

[16] Kurshan, E., Balch, T. and Byrd, D. (2025) 'The agentic regulator: Risks for AI in finance and a proposed agent-based framework for governance'. Available at: https://arxiv.org/abs/2512.11933.

[17] Maurino, A. (2025) 'How data quality affects machine learning models for credit risk assessment'. Available at: https://arxiv.org/abs/2511.10964.

[18] Mupa, M.N., Tafirenyika, S., Nyajeka, M.R., Moyo, T.M. and Zhuwankinyu, E.K. (2025) 'Machine learning in actuarial science: Enhancing predictive models for insurance risk management', Iconic Research and Engineering Journals, 8(8), pp. 493-504.

[19] National Institute of Standards and Technology (2023) Artificial Intelligence Risk Management Framework (AI RMF 1.0). Gaithersburg, MD: NIST.

[20] Nguruve, T., Chitate, I.C., Munjoma, M.B., Makanjera, R., Mutimaamba, V.A., Masunda, M., Mhike, T.I. and Mupa, M.N. (2026) 'Operationalizing AI at scale: Repeatable frameworks for integration, adoption and performance measurement across enterprise and startup environments', World Journal of Advanced Research and Reviews, 30(2), pp. 688-695. doi: 10.30574/wjarr.2026.30.2.1236.

[21] Nhova, M., Mupa, G., Tsveta, L., Dima, J., Chingezi, L., Madzinga, G.T. and Mupa, M.N. (2026) 'Enterprise risk management in health coverage operations: Integrating premium billing controls-first in analytics framework for eligibility-to-invoice integrity', World Journal of Advanced Research and Reviews, 30(2), pp. 1179-1185. doi: 10.30574/wjarr.2026.30.2.1269.

[22] Okpala, I., Golgoon, A. and Kannan, A.R. (2025) 'Agentic AI systems applied to tasks in financial services: Modeling and model risk management crews'. Available at: https://arxiv.org/abs/2502.05439.

[23] Papastergios, V., Ehrlinger, L. and Gounaris, A. (2025) 'Unfolding data quality dimensions in practice: A survey'. Available at: https://arxiv.org/abs/2507.17507.

[24] Ribeiro, M.T., Singh, S. and Guestrin, C. (2016) 'Why should I trust you? Explaining the predictions of any classifier', Proceedings of NAACL-HLT 2016: Demonstrations, pp. 97-101. doi: 10.18653/v1/N16-3020.

[25] Rudin, C. (2019) 'Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead', Nature Machine Intelligence, 1, pp. 206-215. doi: 10.1038/s42256-019-0048-x.

[26] Schneider, J., Abraham, R., Meske, C. and vom Brocke, J. (2020) 'AI governance for businesses'. Available at: https://arxiv.org/abs/2011.10672.

[27] Wand, Y. and Wang, R.Y. (1996) 'Anchoring data quality dimensions in ontological foundations', Communications of the ACM, 39(11), pp. 86-95. doi: 10.1145/240455.240479.

[28] Wang, R.Y. and Strong, D.M. (1996) 'Beyond accuracy: What data quality means to data consumers', Journal of Management Information Systems, 12(4), pp. 5-33. doi: 10.1080/07421222.1996.11518099.

[29] Weber, K., Otto, B. and Österle, H. (2009) 'One size does not fit all: A contingency approach to data governance', ACM Journal of Data and Information Quality, 1(1), pp. 1-27. doi: 10.1145/1515693.1515696.

@article{1719497,
      author = {Liberty Mudzingwa, Tsungai E Tsambatare, Melody Masunda, Munashe Naphtali Mupa},
      title = {Governed AI Agents in Consumer Banking: Data Stewardship and Data Quality Framework for Credit Card and Deposit Analytics},
      journal = {Iconic Research And Engineering Journals},
      year = {2026},
      volume = {10},
      number = {1},
      pages = {377-392},
      issn = {2456-8880},
      url = {https://www.irejournals.com/formatedpaper/1719497.pdf},
      abstract = {Consumer banks and credit unions are moving from conventional dashboards and rules-based automation toward AI agents that can query data warehouses, summarize portfolio signals, monitor exceptions, draft analytical narratives and support frontline and risk teams. The value of these systems depends less on the novelty of the model than on whether the underlying data assets are governed, classified, traceable, timely and fit for regulated decision support. This article develops a practical data-stewardship and data-quality framework for governed AI agents in consumer banking, with emphasis on credit-card and deposit analytics. The framework integrates three bodies of practice: data governance and data-quality scholarship, model-risk and AI-risk governance, and banking-specific expectations around adverse-action explainability, risk-data aggregation and operational resilience. A synthetic consumer-banking test bed is used to demonstrate a risk-weighted data-quality score, business-glossary coverage, data-quality rule monitoring, lineage completeness, access-classification discipline, and AI-agent output controls. Heat-map analysis shows that the highest residual-risk areas are lineage gaps, weak glossary coverage, disputes/chargebacks, credit-bureau attributes and unexplained agent responses. The simulated before-and-after analysis indicates that a governed control layer can raise risk-weighted data-quality scores, improve agent acceptance, reduce escalation and shorten exception aging when controls are implemented as operating mechanisms rather than policy documents. The article contributes a reusable operating model for credit unions, community banks and consumer-banking teams seeking to operationalize AI agents without weakening data accountability, consumer-protection obligations or executive decision reliability.},
      keywords = {AI Agents, Consumer Banking, Data Stewardship, Data Quality, Credit-Card Analytics, Deposit Analytics, Model Risk Management, Business Glossary, Credit Unions, Governed Analytics.},
      month = {July},
      doi = {https://doi.org/10.64388/IREV10I1-1719497}
  }