City Research Online

Abstract

There are very few technologies whose impacts on people’s lives are as widespread and consequential as that of Artificial Intelligence (AI). The domain of finance, broadly speaking, has seen pervasive adoption of AI and has been utilising its transformative potential to change how business is conducted. The primary utilisation of AI has happened inside organisations as firms use it to manage their business operations. Personal finance or consumer-facing financial advisory and management services have seen changes too over the last decade, but their impact has been limited. In the past, financial advisory and wealth management services were only offered by human advisors to those who had substantial financial prowess. AI, through its cost-effectiveness, has democratized the market. AI financial advisors, colloquially called ‘robo-advisors’, have become available to everyone. Research has shown that these AI advisors can aid consumers in making better financial decisions (D’Acunto, Prabhala, and Rossi, 2019). Despite this, consumers have remained hesitant to adopt them. While the adoption or aversion to algorithms has been studied widely in other domains, there is scant literature on the adoption of AI financial products. Although AI-based financial advisory products promise better financial health for people, we are yet to fully understand how to turn them into products and market them in a way that leads to widespread consumer adoption.

In this dissertation, I explore and understand how consumers perceive and adopt AI financial advisory and management products. In doing so, I explore and test the various drivers of adoption and their combined impact on adoption decisions. First, I conduct an interdisciplinary review of the literature on the adoption of AI and related products. I identify gaps in this literature and develop a research agenda. Using this agenda, I conduct in-depth interviews (N=20) to explore consumer perceptions of AI financial services. Through this study, I identify AI-specific attributes which drive product adoption decisions. Subsequently, using three discrete choice experiments (N=3288), I test how consumers trade off these attributes to arrive at adoption decisions. Finally, I discuss the findings of the empirical studies and how they add to our academic and practical understanding of developing AI products and marketing them.

I add to the burgeoning body of literature on the general adoption and aversion of AI and algorithms. My studies show that consumers perceive AI as distinctly different from their human counterparts. They have different evaluative criteria while making adoption decisions, expect different benefits from using them, and do not perceive AI as a direct replacement for human advisors. The findings also indicate that consumers want a specific blend of human and AI expertise when eliciting financial advice. Additionally, I show that issues such as fairness and privacy, among others - often not discussed as regular product attributes - are weighed heavily by consumers when making AI adoption decisions. Thus, I show that the traditional logic of understanding technology product adoption or even pitting humans and AI as direct substitutes may not be the most appropriate way of studying adoption across contexts. I also add to the human-AI augmentation literature by showing that consumers value different attributes of humans and AI and want both involved in different capacities. Consumers expect a new type of experience where firms leverage the differential capabilities of AI and human actors. Based on the findings, I develop practical insights into how firms can build AI advisors that are more likely to be adopted in the marketplace. The dissertation is organized as follows:

Chapter 1 explores how AI technologies are transforming the financial advisory market and how they can add value to consumer financial decision-making. It outlines the gaps in extant research and the scope of this current research.

Chapter 2 systematically reviews the literature on the adoption of AI by consumers and draws out a research agenda.

Chapter 3 addresses the research agenda and explores various drivers of adoption through a qualitative study.

Chapter 4 empirically tests the impacts of the various drivers on eventual product adoption decisions.

Chapter 5 offers a general discussion of the findings of the empirical studies and how they add to our academic and practical understanding of developing AI products and marketing them.

Publication Type:	Thesis (Doctoral)
Subjects:	H Social Sciences > HG Finance Q Science > QA Mathematics > QA76 Computer software
Departments:	School of Policy & Global Affairs > School of Policy & Global Affairs Doctoral Theses Bayes Business School > Bayes Business School Doctoral Theses Bayes Business School > Management Doctoral Theses

Export

Downloads