AI and Financial Inclusion: A Historical Perspective

The ongoing evolution of financial services has consistently altered how credit and financial access are granted. Understanding how AI fits into this landscape requires taking a closer look at three key eras that shaped financial decision-making:

1. The Soft Information Era: Relationship-Based Lending

In the early 20th century, before technological advancements made data widely available, lending decisions were made based on soft information—qualitative judgments and personal relationships. Banks relied heavily on interpersonal interactions to assess creditworthiness. Borrowers’ reputations, community standing, and personal references were essential to determining their ability to repay loans. Lenders conducted thorough interviews and even inspected collateral in person. This highly localized, trust-based model of banking enabled a personalized approach but lacked scalability.

2. The Hard Information Era: The Shift to Quantitative Analysis

By the mid-20th century, lending decisions became more data-driven. Financial statements and quantifiable information such as debt-to-income (DTI) and loan-to-value (LTV) ratios took precedence. This shift marked the beginning of the hard information era, where decisions relied on verifiable and quantifiable data. With the rise of credit bureaus and the introduction of credit scoring models like FICO, lenders increasingly used standardized criteria to assess borrowers. However, this new system excluded many underserved groups, particularly minorities and low-income individuals, who had less access to formal credit histories. This disparity further widened the financial gap.

3. The Financial Technology (Fintech) Era: A Step Towards Inclusivity?

Fast forward to the mid-2000s, the rise of fintech promised a more inclusive financial system. Fintech lenders leveraged machine learning, big data, and digital platforms to streamline lending decisions and potentially close the gap for underserved groups. However, studies showed mixed results. While machine learning-based models like those explored by Amir E. Khandani significantly improved credit risk forecasting, fintech lenders often prioritized high-credit borrowers. Despite advances in technology, financial exclusion persisted, especially for high-risk or minority borrowers.

Is AI the Solution for Financial Inclusion?

As we move into the Generative AI Era, there is growing optimism about the potential of advanced AI technologies to drive financial inclusion. Generative AI, which powers technologies like natural language processing (NLP) and large language models (LLMs), has the ability to process vast amounts of unstructured data (e.g., text, images, audio). This allows lenders to analyze both soft and hard information, creating opportunities for more comprehensive credit assessments. However, the use of AI in this context raises key questions.

1. Can AI Improve Credit Access?

Yes. AI-driven models can process vast amounts of non-traditional data, such as digital footprints from mobile apps and e-commerce platforms. According to research published in the Journal of Financial Stability in 2024, these data sources significantly enhance the predictive power of credit models, allowing lenders to better assess the creditworthiness of borrowers who might lack traditional credit histories.

2. Does Fintech Guarantee Financial Inclusion?

Not entirely. While fintech can accelerate loan approval times and improve risk assessments, studies suggest that many fintech lenders still target profitable borrowers, often excluding those with low access to credit. As documented in the Journal of Financial Economics, fintech lenders often cream-skim the market, prioritizing high-credit borrowers. Although fintech offers potential for inclusion, there is no guarantee that these benefits will reach everyone.

Source: International Banker / Chat GPT