The development of Artificial Intelligence (“AI”) is accelerating rapidly, especially after the extremely successful launch and adoption of ChatGPT. Therefore, we will explore the recent impact of AI on the financial industry, with a specific focus on two financial technology areas: Chatbots and Fraud Detection.
This week we will share the first part of our 2 parts newsletter in which we will briefly describe AI in a nutshell and then review the impact it has brought on the financial industry. We will then introduce Natural Language Programming (“NLP”), which is the AI technology underlying Chatbots and Fraud Detection. Next, we will get into the details of two NLP-based Chatbots, Erica and ChatGPT. We will analyze how they have benefited the financial institutions and their customers. In Part 2 of the newsletter, we will talk about how AI is assisting in the KYC process and Fraud Detection. We will examine how AI has solved challenges and improved the KYC process and introduce Eureka FinTech, a pioneer of AI-powered KYC solutions. The final section of Part 2 will highlight the potential risks and concerns that arise when AI is used in finance, including its limitations, privacy issues, and the risks of discrimination.
This article is also available on LinkedIn.
AI in a Nutshell
AI, or Artificial Intelligence, refers to the development of computer systems that can perform tasks that typically require human intelligence, such as visual perception, speech recognition, decision-making, and natural language understanding. Here are 6 general areas of AI:
AI technology is based on algorithms, statistical models, and machine learning techniques that enable machines to learn from data and improve their performance over time.
The above graph demonstrates that the typical AI process starts with gathering and preprocessing data to ensure that there is an adequate amount of data, the data is of high quality, and it is in a suitable format for analysis. After the data has been pre-processed, machine learning techniques are utilized to train AI models. This model is then packaged and validated on new data. Once the model is validated, it is deployed to perform real-world tasks and monitored regularly to ensure that it is performing as expected. If necessary, the model can be retrainedwith new data to improve its accuracy and reliability.
Role of AI in the Financial industry
The capability of AI to automate complex processes and make predictions based on data can revolutionize the financial industry by increasing efficiency, reducing costs, enhancing quality, raising customer satisfaction levels, and boosting financial inclusion. Thus, it is a win-win for both financial institutions and their customers. Financial institutions can leverage the advanced analytical capacityprovided by AI to gain deeper insights into their customers, enabling them to provide more comprehensive and timely services. As a result, customers using financial services will experience greater automation, leading to increased levels of satisfaction. The utilization of AI technology will benefit all types of customers, including low-income individuals who were previously un-served or under-served by banking services, retail customers, mass affluent, high net-worth individuals, and institutional investors. The adaptability of AI technology allows it to improve access to financial services, promote financial inclusion, offer financial education, and provide tools that can help with optimizing portfolio construction.
With the increased use of AI, the concept of a universal solution has been replaced by personalized financial advice that can be easily accessed with a click, tailored to the unique needs of each client. In order to illustrate, here are some statistics to support and quantify the impact of AI in the financial industry:
1. From The Financial Brand’s research in 2022, AI is estimated to save banks $1 trillion in costs by 2030.
2. From another research by Nadeem Gulzar, Head of Advanced Analytics in Danske Bank, he says that Danske Bank is switching to AI-powered Fraud Detection because it can reduce false positives by 50%.
Natural Language Processing (“NLP”)
Natural Language Processing or NLP is the most widely used AI technology in many financial institutions. NLP is a field of AI that focuses on training computers to learn the patterns of human languages and generate human-like content. NLP is generally used in finance to analyze vast amounts of unstructured data such as news articles, social media posts, and company reports to identify trends, sentiment, and other insights that can facilitate informed decision making. The two most well-known applications of NLP in the financial sector are Chatbots and Fraud Detection.
Chatbots and Fraud Detection applied in financial services will fall into the category of FinTech, or financial technology. By definition, FinTech refers to the use of technology to deliver financial services and products, making them more accessible, efficient, and customer friendly.
Chatbots in Financial Industry
A Chatbot is a communication software using NLP to interact with customers. By analyzing and learning from previous conversations, Chatbots can continuously improve the responses to customer inquiries and requests. Continuous improvement is the most fascinating feature of NLP, and the improvement is achieved through more interaction with customers and gathering feedback into Reinforcement Learning models. So as a customer, if you are continuously using Chatbots for your financial services, you are actually contributing to better development of AI.
Nowadays, Chatbots can assist customers with routine inquiries and transactions, such as fund transfers and account opening, and financial institutions also employ Chatbots for product suggestions and marketing. The banking industry utilizes Chatbots to a great extent. Bank of America’s (“BOA”) Erica is a good example of Chatbots to be put into service on these occasions.
ChatGPT is currently the most powerful Chatbot in the world since its debut in late 2022. According to many credible analysts and articles in the FinTech and finance field, the impact of ChatGPT on the future of customer services can be phenomenal. ChatGPT is built on one of the most powerful Large Language Models (“LLMs”) ever created, called GPT-3.5 (and now has evolved to GPT-4) that can read, summarize, translate, and predict written language. LLM is a type of AI model in the field of NLP, and LLM involves training language models to analyze and understand the patterns of text, either human languages or programming languages, including the specialized vocabulary, grammar, and syntax used in all kinds of documents, daily conversation, and poems, etc. What differentiates ChatGPT from nowadays AI in customer services is that ChatGPT can understand texts and generate human-like responses. Moreover, ChatGPT has an easy-to-use user interface which makes it immensely popular. Here are some specific features of ChatGPT and other LLMs that can generate added values for the customer service in finance:
1. LLMs such as ChatGPT can identify repetitive questions, which eliminates the need for time-consuming manual training, allowing businesses to see meaningful automation and more requests resolved without delay and without the need for human intervention.
2. LLMs and ChatGPT can configure answers with pre-existing data. This means that Chatbots can now understand the data stored in its models and provide customers with identical questions correct and unique responses on its own.
3. LLMs and ChatGPT can collect actionable customer feedback. This is a huge step forward for the service providers as Generative AI is taking a proactive approach to automatically collect and analyze customer feedback in terms of sentiment and propose suggestions to concerns.
4. LLMs and ChatGPT can produce text with a human-like feel. This can engage customers in conversations with Chatbots that closely resemble interactions with human agents.
Based on the above features of ChatGPT or other LLMs, current Chatbots can also be improved. ChatGPT will revolutionize the Chatbots used in the financial industry through its outstanding performance and simple user interface. Chatbots such as BOA's Erica, have already proved to be a highly beneficial tool. The versatility of ChatGPT now allows any financial institution to create its own Chatbot with advanced features tailored to suit its customers' requirements. With ChatGPT's user-friendly interface and developer backend, institutions can effortlessly deploy new ChatGPT-powered applications with minimal IT resources.
BOA’s Chatbot Erica
Before ChatGPT, other Chatbots like BOA’s Erica already provided huge value to the bank’s customer services, even though Erica is less powerful than the latest LLMs. Erica can provide customers with instant assistance and support by automating routine tasks such as balance inquiries, account transfers, and bill payments. In the old days, BOA customers had to set aside time during the day or even make reservations to go to a BOA Branch in person to make these transactions. But now with Erica the Chatbot, customers can do all these transactions with just a few clicks on their mobile app. Erica utilizes NLP to understand BOA customers’ text-based inquiries and answer their questions with helpful information anywhere and at any time. BOA customers can have their problems solved instantly by just typing in their concerns in the conversation box with Erica. As a result, Erica frees up BOA’s human agents to handle more complex requests. The benefits over the traditional customer service methods are clear - 24/7 availability, faster response time, no waiting time and lower operational costs for the banks. To illustrate Erica’s performance, as of 2022, Erica has hit 1 billion user engagement and is growing around 1.5 million per day from then. However, although Erica is a well-developed Chatbot that can perform routine tasks for customers, it still cannot perform complex tasks such as financial planning, which is due to the current limitation of the development of AI, and the required input of human expertise and experience. Yet, BOA’s Erica is still a very good example of how AI has improved the financial industry and provided benefits for both financial institutions and their customers.
More Use Cases and Applications Powered by AI in Financial Institutions
AI, NLP, and Chatbots will continue to revolutionize the financial industry. The use of NLP, one of the AI technologies, in Chatbots allows financial institutions to improve customer services, raise customer satisfaction levels, and increase financial inclusion. BOA's Erica serves as an excellent example of a valuable Chatbot provided by a bank. The highly versatile and flexible ChatGPT platform will revolutionize Chatbots offered in the financial industry, allowing more financial institutions to build their Chatbots effortlessly.
Next week, we will discuss the KYC investigation process and explore how AI can solve traditional KYC challenges. Additionally, we will introduce Eureka FinTech, a FinTech company that integrates AI into its 4D KYC platform. Finally, we will also discuss some of the limitations of AI in the financial industry.
Compiled and edited by Yiwen Chen
Comentarios