In this continuation of our series on artificial intelligence (AI), Advanced Media & Technology of counsel Eyvonne Mallett discusses the intersection of technological innovation and the banking industry. While AI isn’t new to the banking industry, recent advancements have raised concerns about the syndication of personal data, financial predictions, potential bias, and overall privacy and security.
You’ve probably already witnessed the use of AI when you open your banking app. From applying for a credit card to receiving automatic text messages verifying a suspicious charge, you’ve been using technology that the banking industry has long employed, such as machine learning, to streamline the delivery of its financial services. In banking, like in many other industries, controlling the use and collection of data with AI technologies has raised challenges, with the added focus of financial protection for the consumer. In this Q&A, Eyvonne explores the opportunities presented by AI for banks, the legal and ethical challenges and the frameworks needed to leverage AI in a compliant way.
Tell us about your practice and the types of matters you generally handle.
My practice focuses a on broad range of privacy, cybersecurity and contractual matters, with an emphasis on fintech and financial services regulation and compliance. With respect to AI, I am actively supporting clients, including several in the financial and banking industries, on a variety of matters including defining and implementing AI strategies and internal management frameworks, designing AI risk and governance framework controls, advising on licensing and permissions of AI technologies, and monitoring for relevant regulatory updates.
AI and algorithms have long been used in the banking industry for lending decision-making and other automated tasks. What emerging trends impacting banks have you witnessed since the new generation of AI technologies was introduced?
In the world of banking and finance, AI is applicable for a number of potential use cases, such as enhancing customer experiences, improving back-office operations, detecting fraud, managing risk and improving compliance. AI is also used to enable financial institutions to automate repetitive tasks, improve accuracy and speed up processing. Recently, AI-powered chatbots and virtual assistants have been enhanced to provide customers with 24/7 support, speeding up services times and ultimately reducing the need for human interactions.
In the new era of AI technologies, AI is increasingly used for real-time transaction monitoring to quickly assess patterns and identify anomalies, enhanced fraud detection, automated credit checks, hyper-personalized recommendations, customer segmentation, customer behavior and market trend analysis, investment portfolio management, risk management automation, competitor analysis, regulatory compliance (e.g., know your customer, anti-money laundering), and predictive analysis.
Historically, AI assisted in some of these areas, but the enhancements made possible through larger data sets and deep learning models have allowed the technology to become more robust and to make more intricate and accurate predictions when the AI tools are properly trained with reliable, transparent and explainable data inputs.
How is the growing trend of AI adoption transforming traditional banking operations, and which specific areas within the industry are most impacted?
For banks, the biggest draw to AI-driven automation is the operational simplification. In conjunction with a robust AI risk management and governance framework, banks can cost-effectively replace manual tasks, which are often prone to operational failures, with reliable AI tools. This framework ensures a systematic approach to managing risks associated with AI implementation, providing guidelines for responsible and secure use within the financial landscape.
The specific areas within the industry that are most impacted are prediction and recommendation models. These areas leverage AI’s ability to analyze vast amounts of data and to uncover hidden patterns that would not be apparent to a human without a substantial amount of review and time. This has facilitated more accurate and faster predictions, recommendations and decision-making. This pattern recognition has notably been employed in fraud detection and financial forecasting, where about 40% of financial services companies rely primarily on machine learning for both use cases, according to recent market surveys.
What are the major ethical and legal challenges associated with AI in the banking industry?
The major ethical and legal challenges associated with AI in the banking industry are similar to the issues that other industries are grappling with, but with the added requirement to protect against financial risks that can impact the financial stability of consumers and banks.
The use of AI in banking has raised several ethical and legal concerns, including privacy, security, lack of transparency and algorithmic bias. In terms of privacy, AI systems pose challenges concerning how they may process or store personal data without the proper permissions. The security risks presented are related to the potential vulnerability of AI systems to malicious attacks, which can disrupt operations and lead to financial losses. There’s also a concern around lack of transparency due to the difficulty of determining the source of data and how the AI outputs or decisions are made. Lastly, quite possibly the most complex challenge, is how AI systems learn and replicate the biases that may be present in their training data, leading to unfair decision-making and discriminatory outcomes.
Given these challenges, there is the potential for AI decisions to be implicit with bias, inaccurate or, as has been determined in some recent cases, discriminatory. Additionally, when there are weak security measures in place, the technology can be used for nefarious purposes such as money laundering and insider trading, which happens rapidly and may be undetectable because of the speed at which AI processes information. Therefore, banks must have policies, procedures and protocols in place to facilitate the use of AI while mitigating the associated ethical and legal challenges.
Given the regional nuances in regulatory frameworks, how can banks effectively adapt to diverse regulatory frameworks governing AI to ensure compliance and maintain a competitive edge across different jurisdictions?
Banks can adapt to diverse regulatory frameworks by establishing internal AI policies and procedures driven by AI risk management principles that are commonplace in regulation. The primary concerns of the regulators include reliability and biases in the data sources, consumer protection, governance and transparency. For example, the U.S. Executive Order titled “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence” specifically calls out financial services and highlights the importance of data reliability to protect consumers against discrimination, fraud, privacy and cybersecurity risks while maintaining financial stability.
Another example is the EU AI Act, which is not specific to the financial sector but requires financial firms to ensure clear documentation, transparent processing and the establishment of security safeguards when employing AI technologies. Banks can largely ensure compliance with the diverse regulatory frameworks by ensuring policies and procedures address the tenets of privacy, transparency, explainability, auditability, fairness, safety and security, reliability, accountability, and responsibility.
There is much discussion surrounding the concerns and risks associated with AI in banking, but what opportunities does AI present for banks? How can legal counsel support institutions in capitalizing on these benefits while mitigating risks?
The opportunities presented by AI for banks are vast, including improved fraud and risk management and enhanced customer service, automation of mundane tasks, and development of more efficient, highly customized financial strategies. To effectively leverage these opportunities, legal counsel can support institutions by advising on the ethical, legal/regulatory and practical risks and providing guidance on how to mitigate the risks through sustainable, safe and responsible AI use and risk management. This includes education and training as well as the development of an AI risk and governance framework to ensure inclusion of policies and procedures that address the AI risk management principles and regulatory requirements that protect both consumers and banks.
How is Loeb a leader on AI in the banking space?
At Loeb, we have a deep understanding of the industry, its technology, client business needs and the practices that must be employed to balance the ethical, legal and regulatory concerns with regard to the use of AI. We support clients in integrating AI frameworks that enable them to implement compliance and risk controls both internally and externally. Our practical solutions assess risks today while also accounting for forthcoming regulatory requirements. The combination of our knowledge and resources allows us to offer innovative strategies to leverage AI to support the unique business needs of each client.
