AI sweeping Wall Street: Hedge funds deploy ChatGPT, banks engage in an "arms race"

Many people, including billionaire investor Warren Buffett, believe that a passion for adopting complex AI systems is a precursor to future risks.

The artificial intelligence revolution is unfolding in the financial industry.

Hedge fund companies are deploying ChatGPT to handle heavy workload, Deutsche Bank uses artificial intelligence (AI) to scan clients' investment portfolios, and Dutch International Group uses AI to screen potential defaulting parties. Morgan Stanley stated that it is testing the use of AI technology in a safe and controllable environment. At the same time, JPMorgan Chase is widely recruiting AI talent and providing more related recruitment positions than any competitor.

Hand over 'hard work and tiring work' to AI

In fact, AI has long been applied to various works on Wall Street, such as machine learning algorithms for calculating credit risk. At present, Wall Street is exploring the latest popular tools, mainly ChatGPT, in the hope of providing sufficient financial information to enable machines to achieve the ability to reasonably price options, establish investment portfolios, or analyze company news. Financial practitioners need to handle a large amount of financial text data, such as news, reports, ratings, and even deeper level programming work, and AI is gradually taking over.

According to Bloomberg on May 31st, some hedge fund companies have stated that generative AI has been used to handle common tasks such as market research review, basic code writing, and fund performance summary. These "hard work" that once "tormented" junior employees on Wall Street will be digested and completed by AI.

Systematic Quantitative Hedge Fund Campbell& Kevin Cole, CEO of Co, revealed that the company's quantitative analysts use a big language model to summarize internal research reports. Kevin stated that AI has strong capabilities in completing code, editing, and error checking. However, human employees still intervene. The current generative intelligent tools have not yet reached the point where they can change human daily investment methods.

According to a Bloomberg report on June 1st, EigenTechnologies, a company that helps companies such as Goldman Sachs and Dutch International Group carry out AI business, said that the number of inquiries from banks increased fivefold in the first quarter of 2023 compared to the same period last year.

Alexandra Mousavizadeh, CEO and co founder of consulting firm Evident, stated that the release of ChatGPT in November 2022 "made everyone - the board, CEO, and bank leadership - more aware that this is a game changing factor". She described this situation as an "artificial intelligence arms race".

Deutsche Bank is deploying deep learning technology to analyze whether international private banking clients are over investing in a specific asset and match suitable funds, bonds, or stocks for individual clients. On the premise of complying with regulations, real person consultants will pass on AI generated suggestions to customers.

Kirsten Anne Bremke, global head of data solutions at Deutsche Bank International, said, "I really like this way of combining artificial intelligence with human intelligence

JPMorgan Chase has launched a similar plan. According to insiders, the company applied for a patent in May this year for a service similar to ChatGPT, which can help investors choose specific stocks, and the project is currently in its early stages.

BNP Paribas is using chat robots to answer customer questions, using AI to detect and prevent fraud and money laundering. Societe Generale uses AI's computing power to scan for potential misconduct in the capital market.

In April of this year, Morgan Stanley announced that it had applied for a patent for a model aimed at detecting monetary policy direction, using AI technology to classify the Federal Reserve's information as hawkish or dove.

On March 21, Marco Argenti, the chief information officer of Goldman Sachs, a Wall Street investment bank, said that internal developers had begun to use generative AI for programming. This technology is still in its early stages. We will not immediately entrust all important work to AI, but the top priority is to truly try and understand the potential of AI. "Aganti sighed," I have been working in technology for nearly 40 years, and this is one of the biggest disruptions I have ever seen - possibly comparable to the internet, applications, and cloud computing

Risk and safety issues

But this trend still raises concerns about the transparency and effectiveness of financial AI. Many people, including billionaire investor Warren Buffett, believe that a passion for adopting complex AI systems is a precursor to future risks.

Bankers have a fiduciary responsibility not to trade based on unreliable information. Friedman Kaplan Seiler Adelman&, New York; Anne Beaumont, a partner at Robbins Law Firm, stated that as AI applications expand, this is an issue. How do you prove to investors and regulatory authorities that you have fulfilled your responsibilities when you use AI to answer a question without knowing what it is

In April of this year, Samsung Electronics in South Korea was exposed to three incidents where employees used ChatGPT, resulting in the leakage of confidential data. The company's employees uploaded the company's system program code to ChatGPT, requesting AI to help fix errors and improve the program code, and input meeting minutes to ChatGPT, instructing AI to help with key organization, resulting in confidential data such as factory performance and production becoming part of the GPT model training data.

Wang Pengbo, Chief Analyst of Botong Consulting Research Institute, once pointed out that financial practitioners using ChatGPT, even if the user is unconscious, are likely to cause personal information and data leakage.

In addition, both GPT-4 and other large-scale language models suffer from AI hallucinations, occasionally creating and fabricating information out of thin air. If analysts use ChatGPT to generate research reports, the content may seem quite credible, but if false information is sent out in the report, mistakenly becoming a bearish or bullish news for a listed company, it will cause very serious consequences.

In addition, the AI model directly retrieves information training data from the internet, which may contain some copyrighted works. For banks or news agencies that provide financial information, this move is suspected of infringement and may damage the company's reputation.

The financial industry is a highly regulated industry that is particularly sensitive to personal information and related business data. "Wang Pengbo said that when financial institutions use ChatGPT products, they must first conduct a controllable assessment and take preventive measures.

Carol Giovine, a partner at consulting firm McKinsey, stated that when working with banks and insurance companies, they will redesign the risk framework to address risks such as intellectual property considerations, uncertain regulatory environments, and AI "hallucinations".

Campbell& Co's Kohl stated that the company is currently experimenting with using an open-source model internally. Although the processing power is not as strong as ChatGPT, the advantage lies in the ability of the entire system to be deployed and run locally. We must be very careful about the risk of leakage caused by such tools, "Cole said

High development and operational costs

In recent years, the banking industry is no stranger to using technology to gain advantages, and has recruited data scientists, machine learning experts and even astrophysicists. Now, these investments are beginning to bear fruit.

Brian Moynihan, the CEO of Bank of America, stated in April that AI may bring "significant benefits" that help reduce employee numbers, but at the same time, caution should be exercised.

Barclays Bank is still in the "serious research" stage in the financial AI field. C.S. Venkatakrishnan, CEO, said at a meeting that it would take "many years" to implement any AI tool throughout the company.

AIEigenLewis LiuAIHigh development and operational costs

Regarding the current situation in the industry, McKinsey's Giovanni said, "We are currently in a hype cycle, and you can see how quickly the industry is developing. Some banks have begun to realize the conditions required to truly achieve this goal, but many banks are still working hard to understand." At the same time, he pointed out that companies need to identify areas where AI can truly help and develop a roadmap with senior management, Simultaneously train employees and hire more experts.

New Changes in Industry Human Resources

According to the latest data from consulting firm Evident, approximately 40% of vacancies in banks that are most passionate about AI are related to AI recruitment, such as data engineers, quantitative analysts, and governance positions. JPMorgan Chase is at the forefront of the talent race, with data showing that the bank released 3651 AI related recruitment positions globally from February to April, almost twice the number of its competitors Citigroup and Deutsche Bank.

Greg Bond, CEO of ManNumeric, a Boston subsidiary of ManGroup, believes that the 'digital employee' position may be an opportunity for employees who lack technical expertise but are creative and able to ask the right questions. They will have a multiplier effect on your existing research and technical personnel, "Bond said," but ultimately, if we can automate the innovation process itself, that would be a good thing