Blockchain for Ethical and Transparent Generative AI Utilization by Banking and Finance Lawyers

Sachan, Swati ORCID: 0000-0003-0136-0553, Dezem, Vinicius and Fickett, Dale (2024) Blockchain for Ethical and Transparent Generative AI Utilization by Banking and Finance Lawyers In: World Conference on eXplainable Artificial Intelligence (XAI), 2024-7-17 - 2024-7-19, Malta, Valletta.

Text XAI_Conference_Accepted_Version.pdf - Author Accepted Manuscript Available under License Creative Commons Attribution. Download (1MB) | Preview

Abstract

Generative AI tools powered by Large Language Models (LLMs) have attracted significant attention from the banking, finance, legal, and technology sectors due to their ability to generate and articulate coherent human-like text and images. Legal firms have raised ethical concerns regarding LLM's ability to emulate legal reasoning, accountability of erroneous outcomes, and security and privacy of confidential legal data. To address these challenges, this research paper proposes a blockchain-based monitoring framework that ensures the responsible and secure application of Generative AI in drafting legal decisions by utilizing the anonymized output from an existing eXplainable Artificial Intelligence (XAI) algorithm within a law firm, which assists in legal decision-making. The lawyers are expected to comprehend explainable algorithmic decisions expressed in terms of probabilities and feature importance instead of textual explanations. The immutability and decentralization of blockchain technology form the basis of a transparent and tamper-proof record-keeping system. It ensures consistent and tamper-resistant responses by generative AI, which has been used by lawyers in the past. A case study on data security and tort liability claims on banking data breaches is presented to demonstrate the practical application.

Item Type:	Conference Item (Unspecified)
Uncontrolled Keywords:	Generative AI, Explainable AI, Blockchain, Banking, Law
Depositing User:	Symplectic Admin
Date Deposited:	07 Aug 2024 14:53
Last Modified:	23 May 2026 08:51
DOI:	10.1007/978-3-031-63800-8_16
Related Websites:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3183300
Disclaimer:	The University of Liverpool is not responsible for content contained on other websites from links within repository metadata. Please contact us if you notice anything that appears incorrect or inappropriate.