Don't you dare let AI decide your financial strategy?
Author: Hermione He, Ekki Lu, Christy Zha, Sirui Chen
In the corporate world of finance, the presence of Generative AI (GenAI) has been widely felt in multiple areas of the industry. A recent PwC report on global artificial intelligence showed that China and the United States are expected to reap the most abundant reward from the development and adoption of AI, totaling around $10.7 trillion by the year 2030.
Ref: https://www.pwc.com/gx/en/issues/data-and-analytics/publications/artificial-intelligence-study.html
The banking industry could benefit significantly from generative AI. McKinsey estimated that generative AI could increase the industry’s productivity by 2.8 to 4.7% of its annual revenues, equalling around $200 billion to $340 million. Additionally, these generative AI tools could also act as complementary products to enhance employee performance, reduce accumulated technology debt, and deliver personalized content efficiently at a large scale.
Enhancement of worker performance: Generative AI bots can be trained on proprietary knowledge in policies, research, and customer interaction to provide timely support to employees. These bots would be able to consolidate information from public sources and alert customers on semantic queries. For instance, Morgan Stanley is building a GPT-4 AI bot with the combined functions of searching and content creation to help provide their wealth management team with more tailored client information. An European bank has also developed a generative AI-based environmental, social, and governance (ESG) assistant that extracts and organizes unstructured information to generate answers to complex questions. Through the use of generative AI models, companies can save significant costs on back-office operations, especially those associated with accessing information in databases to address customer requests.
Reduction of tech debt: Generative AI tools can aid the software development process in multiple areas. They can generate draft code based on input code and natural language, supporting automatic translations and no- and low-code tools. They can also automatically produce, prioritize, execute, and evaluate various code tests, expediting the testing process and enhancing both coverage and effectiveness. Thirdly, their natural language translation capabilities are valuable in optimizing the integration and migration of older legacy frameworks. Finally, these tools are adept at scrutinizing code to pinpoint defects and inefficiencies in computing, resulting in the creation of more robust and efficient code.
Production of highly personalized content: Generative AI tools have the capability to enhance content creation by utilizing and anlyzing pre-existing documents and datasets. These models could craft customized marketing and sales materials that cater to individual client profiles based on historical data while allowing options for A/B testing. Generative AI also has the potential to write model documentation and search through regulatory information for important updates.
While financial institutions have rapidly embraced AI technology, and the recent emergence of GenAI promises further innovation, the path to progress has its challenges. In this report, we also examine the inherent risks associated with AI in the financial sector, drawing on insights from prominent sources such as the Harvard Business Review and the International Monetary Fund.
Embedded Bias: A Thorn in the Side of Financial Inclusivity
As the financial industry increasingly relies on AI for decision-making, a looming issue takes center stage: embedded bias. Algorithms used in these systems often inherit or exacerbate biases present in training data, leading to potentially discriminatory outcomes. The opaqueness of AI systems makes identifying and rectifying bias a challenging endeavor, casting a shadow over the industry's quest for fairness.
Privacy Concerns: Protecting the Data Fortress
The AI-driven finance sector grapples with an avalanche of sensitive personal data, raising serious concerns about data privacy. Breaches or misuse of this information can have profound consequences. Furthermore, the potential for invasive data collection and profiling poses a significant threat to individuals' privacy rights, underscoring the need for rigorous data protection measures.
Outcome Opaqueness: The Enigma of AI Decision-Making
AI and GenAI systems are celebrated for their complexity, but their decision-making processes often remain inscrutable. This lack of transparency makes it challenging for regulators and customers alike to understand why specific decisions are made. The finance industry faces the potential fallout of this opaqueness, including a lack of accountability and a dwindling trust in the system.
Performance Robustness: The Fine Line Between Brilliance and Vulnerability
The robustness of AI and GenAI systems is another dimension fraught with risk. These systems may shine under specific conditions but exhibit vulnerabilities and errors in unforeseen scenarios. A dependence on these systems without a comprehensive understanding of their limitations threatens financial stability.
Cyberthreats: A New Battleground for Financial Institutions
The financial industry has long been a prime target for cyberattacks. With the rise of AI, new attack vectors and vulnerabilities emerge. AI-powered cybersecurity measures are not invulnerable to exploitation, introducing a novel and evolving risk that financial institutions must confront.
Systemic Risks: The Unforeseen Disturbances of AI Evolution
GenAI's introduction to the financial sector ushers in new systemic risks. The rapid evolution of technology can lead to unexpected market distortions. Additionally, the concentration of AI capabilities in a few large institutions may exacerbate power imbalances, potentially reducing opportunities for smaller players.
While AI has brought tremendous advancements to the financial industry, these technologies come with an array of potential pitfalls. We should remain vigilant and careful while pursuing technological innovation to address these issues in order to ensure the stability and security of the financial sector. Only through careful exploration, comprehensive risk management, and extensive collaboration with AI, we could make use of the potential of AI tools, paving the way for long-term prosperity in the financial domain. As we confront these challenges and opportunities, maintain an open-minded and innovative spirit toward these newest technologies to ensure that the financial industry can continuously adapt to the ever-changing world and embrace the opportunities in the future.
