An Artificial Intelligence Approach to Financial Vulnerability


To understand financial vulnerability and the role of technology in detecting vulnerable customers in banking and how technology affects staff working on this area in financial institutions such as banks.

A large proportion of the adult population in the UK exhibit characteristics of financial vulnerability (FCA 2020). According to FCA Financial Lives Survey in February 2020, about 47% of the adult populationin the UK has some characteristics of vulnerability such as low financial resilience, low capability, poor health,or negative life event, and about 7% are in financial difficulty (FCA 2020).

The large and ongoing increases in household debt in the years prior to the coronavirus pandemic, and the shock of the pandemic have exacerbated these vulnerabilities and led to further indebtedness (Naisbitt 2020). With the steady increase in the use of digital channels to access financial services, and fewer people across the economy being digitally excluded, direct face-to-face consumer interactions with banks is declining making the detection of vulnerable consumers through their behaviours and interactions is making it increasingly difficult for banks to provide appropriate interventions to support their customers. The disconnect between the subjective perception of consumers about their financial resilience and the objective reality is another factor that further complicates and exacerbates the detection of vulnerable customers (O'Connor et al. 2019).

Despite the increasing adoption of Artificial Intelligence and Machine Learning technologies by financial institutions (Ostmann and Dorobantu 2021), the role of AI in the early detection of patterns of behaviour that could identify vulnerable consumers are ignored in the literature. This is due to the disconnect between the social sciences and the necessary sectoral and human perspective this brings, and the technological advancements that AI community pursues. As a result, assessing the impact of recent advancements in relevant areas, such as affective computing and sentiment analysis, are absent from the mainstream financial vulnerability research. Conversely, the AI research is largely oblivious to important human factors involving consumers, employees, and regulatory requirements.

The aim of this study is to respond to this issue by taking a holistic approach to financial vulnerability and bridge the gap between the social science approach and the technological solutions that AI-enabled detection and intervention could bring. We will do so by considering four major areas that create the environment in which consumer vulnerability sits: customers and their experiences, technology and data, staff, and institutional impacts.


In February 2021 FCA published new guidance for firms on the fair treatment of vulnerable customers (FCA 2021). This project contributes to the UK’s aspiration to improve national financial wellbeing by advancing the application and use of state-of-the-art AI technology to detect consumer vulnerability. The findings of the work will inform a key pillar of regulation and advance the approach the finance sector takes to resolving issues in financial vulnerability allowing for the automated monitoring and (re-)assessment of customer needs and their evolution over time.