The Risks, Rewards, and Impact of AI Technology on the Financial Services Industry

, ,

Artificial Intelligence products are dominating the technology sector. OpenAI and its chatbot, ChatGPT, have made headlines for passing standardized testing, generating complex software code, and helping individuals with tax filings. Microsoft Co-Founder Bill Gates said OpenAI’s developments will be “as revolutionary as mobile phones and the Internet.”[1]

ChatGPT has become the fastest platform in the world to reach a user base of 100M only two months after its November 2022 release.[2] For perspective, the previous record-holders, TikTok and Instagram, took 9 and 30 months, respectively, to reach 100M users. With the rapid adoption of ChatGPT, users have started to explore how the tool and other AI solutions can be leveraged in the workplace to improve output quality and efficiency.

How Does ChatGPT Work

ChatGPT is a large language model designed to receive a text-based input, such as a question or prompt, and generate a human-like text output. It generates an iterative response by processing vast amounts of text data, such as articles and online content, to predict the probability of the next word based on the preceding words in the sequence. But it’s not perfect.

  • ChatGPT does not inherently know if a generated response is true or false, and it can generate responses that are misleading or factually incorrect.
  • The majority of the data on which ChatGPT is trained cuts off in September 2021, so the model lacks knowledge of recent current events.

The lack of real-time data is a significant barrier to more widespread commercial adoption in the finance, legal, and consulting sectors, in which both deliverable accuracy and precision are of paramount importance.

ChatGPT Concerns

Wall Street has expressed concerns about data privacy and information security issues surrounding ChatGPT and AI technologies. On March 20th, 2023, a software bug in ChatGPT allowed some users to see the titles of other active users’ chat history and, at times, the first message of a newly created conversation.[3] Data breaches have the potential to violate firms’ client confidentiality agreements and result in a breach of contract. As a result of the aforementioned risks, Bank of America, Goldman Sachs, Citigroup Inc., Deutsche Bank AG, and Wells Fargo & Co. have banned the use of ChatGPT for business purposes.[4]

Still, a February 2023 study from Gartner showed a third of HR leaders do not plan to issue guidance regarding AI chatbots.[5] Leaders may see AI as a passing trend or because they think the responsibility of issuing guidance lies with a different department, such as Legal, Compliance, or IT. However, given the quality control, contractual, and privacy risks, a prudent HR leader should assess potential risks facing their firm and formalize guidance in policies and procedures to ensure the firm is guiding acceptable AI use. See here for a sample corporate policy for Artificial Intelligence Language Models (such as ChatGPT) Usage.

Global AI Reaction

Countries around the world are also taking note of ChatGPT’s potential data privacy issues. As a result of ChatGPT’s bug on March 20th, Italy has banned ChatGPT amid a probe into a suspected breach of Europe’s privacy regulations.[6] Privacy regulators in France and Ireland have contacted Italian regulators to learn more about the basis of the ban. In the United States, the Center for AI and Digital Policy has filed a complaint with the Federal Trade Commission to prevent OpenAI from launching new commercial products based on models past its current iteration, GPT-4. Other AI specialists and industry leaders, including Elon Musk and Steve Wozniak, have called on all AI labs to immediately pause the training of AI systems more powerful than GPT-4 for six months, citing profound risks to society and humanity. They emphasized if labs cannot pause quickly, governments should step in to institute a moratorium.

On April 11th, the Commerce Department issued a formal public request for accountability measures, including whether new AI models should go through a certification process before they are released. The comments, accepted through June 10th, will help advise U.S. policymakers about how to approach AI. On the same day, China’s Cyberspace Administration proposed rules requiring new AI tools from Chinese companies to complete a government security review to ensure they do not generate content that could disrupt social order or undermine state power.

AI Impact on Finance Industry

Despite the skepticism from competitors and regulators, other financial institutions are investing in AI solutions, hoping to gain a competitive advantage. In March, it was reported that Citadel boss Ken Griffin was in talks to secure a business-wide ChatGPT license. Morgan Stanley Wealth Management (MSWM) also announced the division was developing an internal-facing service that leverages OpenAI’s GPT-4 and MSWM’s digital assets and content, allowing Financial Advisors to ask questions and receive digestible, relevant answers with links to internal source documents.[7]

On April 13th, Amazon Web Services (AWS) laid out its generative AI strategy, looking to target corporate customers rather than individual users.[8] AWS said its AI models will be better suited for businesses because they can be trained only on a customer’s data rather than the broader set of web pages that other models like ChatGPT use. Like MSWM’s use of GPT-4 rather than ChatGPT, the AWS approach will offer a safer choice for businesses concerned with protecting private data.

Patomak’s Take

Patomak recognizes the immediate limitations of AI Chatbots and has observed financial institutions incorporating AI restrictions and controls into their policies and procedures, especially to protect against privacy risks. Nevertheless, Patomak also recognizes that as AI models become more accurate and data protections are established, the commercial benefits will expand. In its current state, Patomak notes that ChatGPT may offer time savings benefits for non-confidential daily administrative tasks, such as coordinating travel or scheduling meetings.

The current AI landscape has many similarities to the cryptocurrency landscape in 2021, in which thousands of coins and tokens were created, many of which surged in value, only to crash as people realized the practical use cases were limited or non-existent. In 2023, hundreds of AI tools have been launched with lofty promises of improving productivity, innovation, and decision-making. Despite the promises, market participants should consider if AI solutions are well-suited for their business needs, or if a simpler and more cost-effective solution exists. It seems likely that once the initial AI shine wears off, the market will pick the winners and losers in the space.

[1] Bill Gates, The Age of AI Has Begun, GatesNotes, March 21, 2023,

[2] Krystal Hu, ChatGPT sets record for fastest-growing user base: analyst note, Reuters, February 2, 2023,

[3] OpenAI Staff, March 20 ChatGPT outage: Here’s what happened, OpenAI Blog, March 24, 2023,

[4] Gabriela Mello, William Shaw, and Hannah Levitt, Wall Street Banks Are Cracking Down on AI-Powered ChatGPT, Bloomberg, February 24, 2023,

[5] Jo Constantz, Nearly Half of Firms Are Drafting Policies on ChatGPT Use, Bloomberg, March 20, 2023,

[6] Supantha Mukherjee, Elvira Pollina, and Rachel More, Italy’s ChatGPT ban attracts EU privacy regulators, Reuters, April 3, 2023,

[7] Morgan Stanley Staff, Morgan Stanley Wealth Management Announces Key Milestone in Innovation Journey with OpenAI, Morgan Stanley, March 14, 2023,

[8] Tom Dotan, Amazon Joins Microsoft, Google in AI Race Spurred by ChatGPT, The Wall Street Journal, April 13, 2023,