Legacy Banks at the Crossroads in the AI Revolution

When one hears the word “legacy” the first thought most likely would be something related to the past. It usually carries a positive overtone. If said in the context of a bank, one would think the bank is no longer open. The “legacy” business which will be discussed describes a longstanding, community-serving businesses that are valuable cultural assets to the city (Legacy Business Definition | Law Insider). The city of Los Angeles describes a legacy business as operating ten or more years (https://ewdd.lacity.gov/). For the sake of this article, we will add even more nuance to the definition, adding one hundred or more years in operation and in the sector of banking.

Why focus on legacy banks? Well, I happen to work for one, but as I reflect over my career, I have worked for two legacy companies. My current role at a 108-year-old bank in Des Moines Iowa and previously at a 130-year-old retailer. It has me reflecting on some of the unique aspects of these legacy businesses.

Origins are family owned and community focused

Many legacy businesses start small and mainly focused serving the local community in which it was founded. This is true with my current employer as well as former. They are typically founded to meet the needs of a specific or growing need in their local community. This community focus over time builds trust and loyalty among customers and gradually becomes a staple of said community.

Weathered many turbulent times

Based on our definition of legacy, companies who last over a century have weathered, depressions, world recessions, world wars, pandemics, and other perils. Often what keeps legacy organizations in business is the original premise of their founding, which is the focus on their values and the origins.

Let us zoom in on legacy banks. What I see as some of the common challenges that face legacy banks is adopting cutting edge technologies, while remaining competitive in a highly regulated and fiercely competitive market segment. Artificial intelligence and its many applications are one of them that can be a shock to the system for legacy banks. The technology is evolving a pace that makes it hard to keep up with not only its many applications, but also its many pitfalls if not done correctly. This makes banking executives nervous because as fiduciaries, the safeguarding of assets and customer information is paramount and top of mind. Anything new and not fully understood, is often viewed as a threat to that responsibility.

Technology characteristics of legacy banks include:

• Aging and outdated technology infrastructure
• Massive data sets with little insight to the data
• Outdated customer facing technology
• Improperly scaled operations support units

Culture

Peter Drucker an Austrian-American management consultant, and who is widely credited through his Management Theory, and many books such as The Practice of Management is quoted as saying, “culture eats strategy for breakfast.” While the exact quote is disputed it is attributed to Drucker being one the

preeminent experts of modern corporate structures. Drucker was on to something. As anything new is introduced to a legacy company, the first obstacle that it will meet is the company culture. If the company culture is resistant to change, most new and innovative ideas will die on the vine. To counter this, relating to AI, the culture of the company must have change champions sprinkled throughout the different levels of company. These champions are responsible for socializing the technology, educating of its benefits, showing the value proposition, and take on any preconceived ideas, fear, and other objections that will arise. If you reflect on banking, all legacy banks today would have gone through some kind of technology transformation. From paper-based ledgers to spreadsheets, from paper checks to electronic checks, from manual credit card readers to tap to pay and virtual cards. All of these represent technological advances that without it, banks would not be able to compete the raise of aggressive FinTech’s offering sophisticated niche solutions. It will be the same with artificial intelligence. The efficiencies to be gained, insights to be begotten, is going to be essential in competing in an ever- evolving technological landscape. The culture must now incorporate AI and its many applications to cultural customs and thought processes.

Baby Steps to Transformation

Now that we have unpacked what legacy banks are and the challenges it will face with AI, let’s look at how a legacy bank can properly transform itself. For the purpose of this article, we will not prescribe any particular platform, or software package but instead focus on what the software can do.

• Automation first-The gateway into more advanced AI approaches is Robotic Process Automation (RPA). RPA is taking a predictable, structured, linear, rules-based process and programming a “Bot”, or automation software to mimic the actions of the human. The target of these automations should be anything that is manual in nature and does not provide any value-add to the business. That’s not to say the process isn’t needed, it is to say the human can be freed of this manual work and shift their focus to more value-add work. Most manual processes are fraught with errors. Freeing the human from these will increase moral and tap into other talents the human possesses. The most effective way to introduce this technology is to find the quickest wins that can be demonstrated to show impact. Quick wins allow you to also show business units what is possible. This will usher in creative thinking in how the technology can be implemented. Another possible side effect of freeing the human of mundane and manual work, is that innovation and innovative thinking tends to happen in that time that is free to explore and think outside of the processes that consume the employee’s time.
• Intelligent Document Processing and other smarter automations-Once the basics of automation is now in production and clear value is observed, it’s time to start introducing more advanced automation. The next logical area of focus is Intelligent Document Processing (IDP). Manual paper intensive processes are the norm in banking, especially for the back-office support areas. IDP allows for the reading, routing, and automated processing of a myriad of document sources. Innovation is born when everyone is liberated from mundane, manual non-value-add work.
• Generative AI and Machine Learning Approaches-Generative AI generally describes the use of AI coupled with some type of Large Language Model (LLM’s), or Small Language Model (SLM’s) that can create new content and ideas, including conversations, stories, etc. It can learn human language, programming languages, art, chemistry, biology, or any complex subject matter and continues to learn over time (https://aws.amazon.com/what-is/generative-ai/). Once the first two have been implemented the move to more sophisticated approaches can be a natural transition. Specific use cases for GenAI can be internal chat bots that allows employees to use prompts to chat with specific data sets such as employee benefits information, company policies, etc. Additional capabilities include the ability for organizations to interact with their customer data and find valuable insights into customer patterns, potential market opportunities, and gain a 360-degree view of their customer. As this approach matures, attention can be turned to public facing GenAI solutions for customers.
• Establishment of proper governance and oversight-Music to the ears of board members and bank leadership is hearing that all AI activities have a proper governance structure and operational framework to ensure the technology is kept within the bounds of all legal and regulatory requirements. The establishment of a Center of Excellence (CoE) for the proper monitoring and controls with representatives from audit, IT, risk, data governance, legal, marketing, human resources, etc. will ensure all the proper lenses are looking at and monitoring their domains for adherence to all legal, and regulatory requirements.

What has been discussed here is not an easy proposition. On one hand, legacy banks have had to transform themselves over time technologically. Gone are the days of manual credit card readers, large paper ledgers, paper checks for the more efficient electronic versions. Innovative technology is not new to banks. The difference with AI is the transformation is at a faster pace, fraught with pitfalls if not done properly, but with an upside that can play a key role in ensuring the legacy bank continues the legacy well into the future.