Applications of Generative AI in the Financial Sector

Introduction

The finance industry is the cornerstone of any country’s development, as it drives economic growth by facilitating efficient transactions and credit availability. The ease with which transactions occur and credit is availed determines the fluidity of markets. It also encourages investments and fosters innovation. Additionally, the increasing demand for financial services makes it critical to constantly update the technology involved in these services. And the latest trend in this regard is the use of generative AI (GenAI) in the financial sector.

The McKinsey Global Institute (MGI) estimates that across the global banking sector, GenAI could add between $200 billion and $340 billion in value annually, or 2.8 to 4.7 percent of total industry revenues, mainly through increased productivity.

This may make you wonder if the finance industry is switching from traditional AI to generative AI. Well, let’s explore some applications of generative AI in the finance industry.

Overview

• Explore the various uses of generative AI in the financial sector.
• Understand how GenAI models in financial services help with scenario analysis, fraud detection, etc.
• Learn how companies like PayPal, BlackRock, and Mastercard use generative AI in their workflows to enhance productivity and hyper personalize communications.

Synthetic Data Generation for Scenario Analysis and Fraud Detection

Financial institutions have a vast amount of customer data. So it is fair to assume that training a model would be easy peasy. But it is easier said than done.

One problem the financial institutes face while training models for fraud detection or scenario analysis is the lack of enough instances of such incidents. Imagine a scenario where you have a dataset of millions of transactions out of which only 100 transactions are fraudulent. It is very likely that the fraud detection model fails to predict fraudulent instances due to class imbalance in the training dataset.

Similarly what about scenarios that have never happened before in that financial services company? Any financial service would want its model to predict a disastrous financial scenario for which it may not be trained. But since the existing model is not trained on such extreme scenarios, even scenario analysis seems like a far-fetched dream. This is where synthetic data comes into play.

You can generate synthetic data to train your models for scenarios that have never occurred before. These could range from the most significant financial frauds to how the bank will perform when a macroeconomic disaster strikes. Hence, the proper adoption of GenAI into financial services can be a game changer for any economy.

A prominent example of a financial service adopting this is Mastercard, which is using synthetic data to improve its fraud detection model.

Also Read: Visa’s Cutting-Edge AI Shields Credit Card Users Against Cyber Threats

Productivity Enhancement with GenAI Integration

One of the pressing pain points of financial services is delivering results as quickly as possible. Thus, the integration of generative AI into their workflows is imperative to bring maximum efficiency into the system.

PayPal’s GenAI platform, Cosmos.AI, powers AI-driven operations, enabling tasks like fraud detection and personalized customer service. Using techniques like Retrieval-Augmented Generation (RAG) and semantic caching, Cosmos.AI enhances chatbot functionality, improving PayPal’s workflow efficiency and reducing operational costs.

Another instance where GenAI integration boosted productivity is the lending tech giant Zest AI’s LuLu. It helps lending institutions analyze portfolio performance, access industry insights, and optimize decisions with natural language prompts.

LuLu allows lenders to ask questions like “How does my approval rate look over time?” and receive instant, data-driven responses, enhancing decision-making and agility.

Hyper Personalised Communication for Customer Satisfaction

Imagine you are applying for a home loan. Here’s a rough set of steps you will be following through the process:

1. You contact the bank and request a home loan.
2. They send you the details of the loan, c
3. You fill out the form (you may think, why does the bank not fill in the details when they already have your data) and mail it to the bank.
4. Then, due process is followed before approving or rejecting the loan application.

Sounds tedious, right?

Now, let’s imagine a scenario where this communication is taken over by a generative AI tool powered by LLM. This LLM is fine-tuned to understand the financial rules and regulations of the geography. It also has access to the bank’s relevant databases and the home loan document policies. Here’s potentially how the process would flow:

1. You can apply for the home loan on the relevant page.
2. The LLM checks your eligibility for the loan based on your income, credit score, and other information.
3. If you are eligible for the loan, it automatically fills the form with the available details. If the tool does not have a particular information, it sends a message asking you to fill that field and finally double-check the pre-filled form.
4. If you are not eligible for the loan, a personalized message will be sent to you, citing the reason for rejection.

Note that all of this happens within minutes! Yes, a fraction of the time compared to the traditional method.

Financial Institutions like DBS, Standard Chartered, and NCR Voyix have already started using GenAI for this process by integrating Kasisto. This leading digital experience platform helps them fasten up communication and other processes involving organization-bank interaction. Additionally, you can also get answers to questions like, “How much did I spend eating out last month?” without creating those darn Excel sheets. Needless to say, it will be an exciting time ahead, tracking your expenses and getting a reality check on your spending.

Asset/Portfolio Management with Generative AI

Asset management is another effective use case of generative AI in finance. It deals with maximizing portfolio value by buying or selling assets like stocks, bonds, real estate, etc. while minimizing risks according to the client’s goals and time horizon.

Earlier, some financial services were using business intelligence tools like PowerBi and Tableau to prepare charts, get information on portfolio performance, and assess risk. It was a time-consuming process as a lot of manual work had to be done, and the job role was limited to people who were pros at using such tools.

However, with GenAI, you can simply write a prompt and get the information. eFront (a part of BlackRock) has launched its copilot, which enhances decision-making and data analysis for private market investors. This portfolio querying tool improves efficiency by automating data workflows and providing real-time insights, eliminating the need for manual report generation.

One can ask eFront copilot, “What is my exposure to the manufacturing sector?” or “Group the data by country” with simple prompting, and voila!! You will get your output.

Conclusion

Generative AI is the holy grail, giving a new life to the finance industry. From enhancing efficiency, decision-making, and customer experiences to synthetic data generation for fraud detection, hyper personalized customer communication, and real-time portfolio management, GenAI is re-writing traditional processes. This adoption comes with the benefits of using advanced AI capabilities to stay competitive, reduce costs, and deliver hyper personalized services. It will be exciting to see how the adoption progresses as the world of generative AI moves forward.

Also Read: Applications of Machine Learning and AI in Banking and Finance in 2024

Frequently Asked Questions

My name is Abhiraj. I am currently a manager for the Instruction Design team at Analytics Vidhya. My interests include badminton, voracious reading, and meeting new people. On a daily basis I love learning new things and spreading my knowledge.