Wall Street Bans ChatGPT | March 2023

March 29, 2023

Geoffrey A. Hirt

Artificial intelligence (AI) relates to machine (computer) learning that is able to perform activities and tasks that usually require human intelligence such as decisions, visual perception, and speech recognition. In short, it makes computers act like humans. Recently, the spotlight is on generative AI. Wall Street is putting restrictions on generative AI, prohibiting employees from using the popular tool ChatGPT.

What is generative AI?

Generative AI can produce content including text, images, audio, and more. AI chatbots (such as ChatGPT), image generators (such as DALL-E), and other generative AI tools are changing business operations. These tools can be used to write copy, process customer data, answer common questions, automate routine tasks, translate documents, and more.

What is ChatGPT?

ChatGPT is an AI-powered chatbot developed by OpenAI, an AI research and deployment company. Since it was launched in late 2022, many organizations have scrambled to address the technology. Some companies are embracing it while others are concerned about its use.

When we asked ChatGPT how employees in the banking industry could use its services, it said:

Customer service: “Employees in banking and financial services can use ChatGPT as a virtual assistant to assist customers in real-time. ChatGPT can answer frequently asked questions about banking products and services, such as account opening, credit/debit card queries, loan applications, and more.”

Risk management: “Employees in the banking and financial services industry can use ChatGPT to monitor risk and identify potential fraud. ChatGPT can detect unusual patterns in transactions, flagging suspicious activity and alerting employees to investigate further.”

Compliance: “ChatGPT can be used to ensure that employees comply with regulatory requirements. For instance, employees can query ChatGPT to check whether a certain activity is allowed under applicable laws and regulations.”

Training and development: “ChatGPT can provide guidance on best practices, policies, and procedures. Moreover, ChatGPT can simulate real-world scenarios to help employees practice decision-making and problem-solving skills.”

Market research: “ChatGPT can be used to gather insights on customer behavior, preferences, and opinions. Employees can query ChatGPT to conduct surveys, collect feedback, and analyze sentiment.”

Wall Street cracks down

Bank of America, Citigroup, Deutsche Bank, Goldman Sachs, Wells Fargo, and others have banned or restricted ChatGPT for business use. Many of the restrictions were automatic since each organization has established guidelines for implementing third-party software.

There are many reasons to be cautious when it comes to generative AI. First, the technology is relatively new, so businesses must determine how it can be used ethically, safely, and effectively. Second, as of this writing, ChatGPT is not transparent about how it synthesizes and formulates answers to prompts. This could lead to plagiarism. Other programs, such as Microsoft Bing’s chatbot, address this by using footnotes to cite sources. Third, generative AI tools may generate incorrect or incomplete information. Though these tools offer some benefits, employers may fear that their employees will misuse them or over-rely on them instead of thinking for themselves.

In the banking and financial services industry, trust is paramount. Since there are many unknowns with how ChatGPT sources the information for its outputs, employees using ChatGPT to create content could be putting their bank’s reputation at risk. For instance, financial advisors are often hired for their professional expertise and insights. Their authority could be undermined if it was discovered their monthly communications with clients were written by a chatbot.

The future of generative AI in finance

There is also some concern that AI could replace financial advisors or challenge services offered by banks. Right now, the functionality of ChatGPT is somewhat limited, and the information it uses is not updated in real-time, but it’s possible that advanced AI could play an increasingly important role in the industry.

While AI can answer basic questions like account balance or payment due, people in the financial industry rely on a personal relationship with their financial advisors. For example, an account executive knows their client and often their families. Each client has a different risk profile and set of financial goals. An account executive may have clients of all different ages, wealth, and risk tolerance. In the end, the question might be whether a chatbot can replace the relationship between a financial client and their account executive. How would you feel about having a financial relationship with a chatbot?

In the Classroom

This article can be used to discuss the future of banking (Chapter 15: Money and the Financial System).

Discussion Questions