Tackling the hidden artificial intelligence bias in the financial sector: A literature review of the impacts and strategies for mitigation
Year: 2023
Type: dissertation
Abstract: Artificial Intelligence (AI) has evolved rapidly in recent years due to improvements in computational power and the exponential growth in the available data. These improvements have made AI more practical and precise, and it has been implemented at an increasing pace in different sectors worldwide. The financial industry has been one of the most prominent fields for AI adoption due to the sector’s diversity. Consequently, AI is nowadays an essential part of the functioning of financial institutions, FinTech companies, and financial markets. The benefits often blind the operators from the inherent threats of AI, such as AI bias. AI bias might increase inequality and harmfully impact individuals, organizations, societies, and economies. Thus, it should be in everyone’s interest to tackle the issue. The current literature seems to lack a coherent understanding of the harmful effects of AI bias and strategies to mitigate the problem. Therefore, this study aims to investigate AI biases’ different impacts and identify mitigation strategies. This thesis conducted a literature review and found that AI bias has negative impacts on both micro- and macro-levels. The critical nature of the financial sector indicated that these effects are severe and long-lasting. Similarly, mitigation strategies varied from local micro-level actions to global macro-level efforts. Mitigating AI bias will likely require a combination of different strategies and take some time, as the issue is deeply rooted and widely spread.
Source: Aaltodoc (Aalto University)
Author: Paloheimo, Oona
Institution: Aalto University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7025113909
