MCP-ORIENTED GENERATIVE AI ASSISTANTS: TRUSTWORTHY AND POLICY-DRIVEN ARCHITECTURE FOR FINANCIAL SERVICES
DOI:
https://doi.org/10.30890/2567-5273.2025-41-03-055Keywords:
generative assistants, large language models (LLM), Model Context Protocol (MCP), Retrieval-Augmented Generation (RAG), semantic search, compliance, AML/KYC, operator copilot, guardrails, Model Risk Management (MRM), PII, latency SLO, p95/p99Abstract
Financial institutions have accumulated vast corpora of regulatory policies, procedures, case notes, and operational documents, creating an overwhelming cognitive burden for compliance officers, risk managers, and operations teams. Traditional informationReferences
Pandey, Varun. (2025). Agentic AI with retrieval-augmented generation for automated compliance assistance in finance. International Journal of Science and Research Archive. 15. 1620-1631. 10.30574/ijsra.2025.15.2.1522.
Iaroshev, Ivan & Pillai, Ramalingam & Vaglietti, Leandro & Hanne, Thomas. (2024). Evaluating Retrieval-Augmented Generation Models for Financial Report Question and Answering. Applied Sciences. 14. 9318. 10.3390/app14209318.
Zhao, Suifeng & Jin, Zhuoran & Li, Sujian & Gao, Jun. (2025). FinRAGBench-V: A Benchmark for Multimodal RAG with Visual Citation in the Financial Domain. 10.48550/arXiv.2505.17471.
Gan, Aoran & Yu, Hao & Zhang, Kai & Liu, Qi & Yan, Wenyu & Huang, Zhenya & Tong, Shiwei & Hu, Guoping. (2025). Retrieval Augmented Generation Evaluation in the Era of Large Language Models. 10.48550/arXiv.2504.14891.
Konda, Snehansh. (2024). The Integration of Large Language Models in Financial Services: From Fraud Detection to Generative AI Applications. International Journal of Scientific Research in Computer Science, Engineering and Information Technology. 10. 1652-1665. 10.32628/CSEIT241061208.
Harrison, William & Pum, Mengkorn & Vaithianathan, Muthukumaran. (2024). AI for Anti-Money Laundering (AML) and Know Your Customer (KYC) Compliance.
Suryavanshi, Kameshbhai & ks, Sendhil. (2025). AI-Driven Financial Risk Management with LLM Agents and Adaptive Learning.
Lin, Huawei & Lao, Yingjie & Geng, Tong & Yu, Tan & Zhao, Weijie. (2025). UniGuardian: A Unified Defense for Detecting Prompt Injection, Backdoor Attacks and Adversarial Attacks in Large Language Models. 10.48550/arXiv.2502.13141.
Yan, Jun & Yadav, Vikas & Li, Shiyang & Chen, Lichang & Tang, Zheng & Wang, Hai & Srinivasan, Vijay & Ren, Xiang & Jin, Hongxia. (2023). Virtual Prompt Injection for Instruction-Tuned Large Language Models. 10.48550/arXiv.2307.16888.
Garza, Leon & Kotal, Anantaa & Piplai, Aritran & Elluri, Lavanya & Das, Kumar & Chadha, Aman. (2025). PRvL: Quantifying the Capabilities and Risks of Large Language Models for PII Redaction. 10.13140/RG.2.2.21858.64967.
Narayanam, Deepak & Singh, Trilok. (2025). AI-Powered Regulatory Compliance Exploring the Role of LLMs in Automating AML Documentation and Reporting Workflows. 10.13140/RG.2.2.31207.36004.
Berger, Armin & Hillebrand, Lars & Leonhard, David & Deußer, Tobias & Bell, Thiago & Dilmaghani, Tim & Kliem, Bernd & Loitz, Rudiger & Bauckhage, Christian & Sifa, Rafet. (2023). Towards Automated Regulatory Compliance Verification in Financial Auditing with Large Language Models. 4626-4635. 10.1109/BigData59044.2023.10386518.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2025 Authors
This work is licensed under a Creative Commons Attribution 4.0 International License.
