Background Information -

What is Generative AI? Agentic AI?

Since the introduction of ChatGPT in early 2023, Large Language Models have become one of the most important technological shifts globally, and the growing influence of LLMs in the finance sector is a shift that directly challenges how entry-level work is structured in the industry. LLMs—often referred to as generative AI—is defined in “(Generative) AI in Financial Economics” as systems “trained using self-supervised learning and capable of producing outputs that closely resemble human-generated content.” (Mo and Ooyang 512) This type of AI not only has the ability to generate text, but also the capacity to replicate perceptive and analytical tasks. This capability is incredibly useful in financial services due to an AI’s ability to process and generate information at scale, and especially in areas such as private credit, depending heavily on analyzing large amounts of data. Mo and Ooyang in their article further explains the importance of LLMs, claiming that AI has “begun to reshape the financial system by significantly enhancing its information processing capabilities – the core function through which it allocates resources, manages risk, and supports economic coordination” (Mo and Ouyang 510). The economists’ use of the phrase “core function” implies that improvements in information processing from generative AI will not only improve efficiency, but rather become a fundamental transformation for making financial decisions. Because the finance sector depends heavily on analyzing and interpreting information, LLMs will inevitably affect roles heavily based on processing information.

The rapid evolution of LLMs with greater capabilities and autonomy further raises questions about the future roles of entry-level workers. Aside from generative AI, LLMs have also developed into what is known as agentic AI, as the article, “Measuring agentic AI adoption and control frameworks in finance” described it a system “that operate as goal-directed agents, integrate with enterprise tools and workflows, and exercise some degree of delegated action rights…under specified constraints” (Mustafa and Aysan 81). The concept of “delegated action rights” is especially significant, because it suggests that AI is no longer limited to assisting human decision-making, but can begin to act on behalf of users within structured systems. Agentic AI may reduce the need for workers to perform these foundational tasks, and therefore will change the roles demanded from firms.

What is Private Credit?

As artificial intelligence becomes increasingly capable of outperforming traditional human tasks, the finance sector has become one of the industries most exposed to technological transformation. Private Credit—a form of debt financing, involving non-bank lenders providing loans directly to companies outside of public markets—has become an increasingly powerful sector of finance and the banking industry. Masaaki Yoshimori in his research “Artificial intelligence, private credit, and the formation of financial bubbles: Evidence from U.S. institutional lending dynamics from shadow banking” explains that private credit went from “a niche asset class into a structural pillar of modern financial intermediation,” and further states that it has already “blurred the boundary between market-based and relationship-based finance.” (Yoshimori 1). This suggests that private credit now plays a central role in how capital flows through the economy, reinforcing its importance in finance. However, since private credit is heavily reliant on data analysis, forecasting, and risk evaluation, it will be especially vulnerable to AI exposure. This exposure does not necessarily mean AI will negatively impact private credit, since LLMs are extremely capable of improving efficiency, though many of the responsibilities of entry-level employees overlap closely with the capabilities of LLMs. In the article “Algorithmic governance in banking: a comparative analysis of risk-based and accountability-oriented oversight,” researchers found that AI systems are able to “mediate credit scoring, market surveillance,anti-money-laundering processes, and prudential oversight.” (García-Llorente and Olmeda 18) The responsibilities mentioned are typically “grunt work” assigned to entry-level workers, with now being able to be “mediated” through AI suggests that LLMs are increasingly becoming an active participant in finance and banking.

Financial Labor Market

The relationship between AI and the financial labor market provides more information to understand how LLMs might affect new graduates. Anthropic’s insight on “Labor market impacts of AI” by Massenkoff & McCrory provides a measurement for AI displacement risk called observed exposure, calculated using theoretical LLM capabilities and real world usage data. Their findings further suggest that “occupations with higher observed exposure are projected to… grow less by 2034,” while adding that there is “suggestive evidence that hiring of younger workers has slowed in exposed occupations.” (Massenkoff & McCrory 2) This evidence shows that AI will not eliminate but rather slow job occupation growth in fields where AI can perform similar tasks. The researchers further explain that “financial analysts are among the most exposed” occupations to AI integration, adding that “there’s tentative evidence that hiring into those professions has slowed slightly for workers aged 22-25.” (Massenkoff and McCrory 14) In positions such as private credit where AI exposure is more likely, there will be suggestions to reduce hiring for entry roles and increase expectations for the skills required to enter the field, causing new graduates to face more competition.