AI and FinTech: The ultimate power duo that’s changing the way we navigate the world of finance

 

By Anastasio Scalisi, CTO of Revolv3. 

It’s 2023 and Artificial Intelligence is changing the face of the financial services industry. On a very fundamental level, by enabling organizations to access and utilize relevant data hitherto inaccessible to them, AI is equipping them to make informed and efficient decisions every day. The phenomenal combination of data analytics and AI are being widely applied across the financial industry to enhance user experience, detect fraud, forecast revenue, predict stock prices, monitor risk, and more. In short, what was once considered nice to have has become the beating heart of the industry, propelling its growth to exponential levels, also fueled by the customers’ growing willingness to use digital channels even for complex interactions. 

AI is impacting the financial services landscape in every conceivable way possible. What’s happening and what’s in store for the future? Let’s start with something we interact with way more often than we’d like to admit- payments. 

AI in payments 

It’s easy to fall into an overwhelming rabbit hole of the possibilities, pros, and cons of Artificial Intelligence, whereas the areas that may benefit the most from such innovations might be hiding in plain sight. That’s why it is important to figure out exactly where to implement AI solutions so that existing processes can be enhanced and made as frictionless as possible. For any industry and any niche within, the first step could be as simple as asking ourselves what the most pressing issues that we face in our routine workflow are and whether AI can do a better job at it. Consider false credit card declines for instance, which is most commonly caused by banks mistakenly flagging a valid transaction as fraudulent. What happens if we place AI-powered algorithms to prevent this? By identifying and flagging irregularities with accuracy, these algorithms can drastically cut down on false alarms that are frequent causes of customer frustration and in many cases, a subsequent churn. 

AI helps optimize the process of dynamic routing by analyzing data in real time to identify the most suitable payment channels for each transaction based on factors such as payment amount, location, currency, and merchant configurations. AI can also monitor and predict network congestion and adjust routing in real-time to ensure optimal transaction processing.  When an optimal route results in certain declines, AI uses the transactional data and available options to retry across processors (if enabled) in real time before the response is even sent back to the merchant. 

Since we’re on the topic of irregularities and unusual patterns in transactions, it’s worth mentioning that the potential of AI in fraud detection and prevention of financial crimes is certainly a game-changer for the industry. Given the exceedingly large number of digital transactions happening every hour of the day, cybersecurity also needs to be ahead, at every stage. The best thing about AI and ML is that the algorithms are always learning to get better at what they do. When fraudsters get smart, institutions need to get smarter. For example, Mastercard’s fraud-detection machine learning system uses supervised learning to identify established fraud patterns and unsupervised learning to identify emerging fraud patterns in rea time. In addition to real-time data on card usage such as what they are trying to buy, where they are trying to buy it, and what else they bought the same day,  the machine learning algorithms also examine various aspects such as a cardholder’s purchasing behavior, location, and travel patterns, with every transaction. Each transaction is evaluated in terms of the rules that relate to what constitutes a valid transaction and what a fraudulent transaction looks like. 

Reducing costs, Growing revenue 

Accenture forecasts financial services will benefit $1.2T in additional GVA in 2035 from AI. There is a palpable sense of optimism in the industry too- a survey by Nvidia found that nearly half of the respondents said that AI will help increase annual revenue for their organization by at least 10% and more than a third noted that it will also help decrease annual costs by at least 10%. 

It’s been highlighted that AI has enhanced business operations, especially when it comes to improving customer experience, creating operational efficiencies, and reducing total cost of ownership. For financial institutions, automation of processes like financial document analysis and claims processing frees up a considerable amount of resources in terms of time, expenses, and human input, thanks to computer vision and natural language processing. Through efficiencies generated by higher automation, organizations can reduce error rates and utilize resources more optimally. Uncovering previously unrealized opportunities, powered by a renewed ability to process and generate insights from extensive troves of data means they are opening new, lucrative streams of revenue.  

Can AI help bring the human touch to customer experiences?

From onboarding and handling first-level customer inquiries to offering hyper-personalized recommendations, AI has been impacting every aspect of customer experience for FinTechs. Would we have imagined the possibility of robo-advisers in the past? It’s undeniable that leveraging AI to enhance self-service options will make the customer’s search for information effortless and the experience more interactive- think chatbots and conversational AI getting impressive by the day. Empathy, however, remains an attribute that no digital experience can do without. 

According to Accenture, empathetic banking leaders leverage their ability to gather and use data insights to engage with their customers by offering them channel choices based on their emotional state and financial need instead of forcing them through a particular touchpoint. By constantly collecting data that helps them anticipate the appropriate kind of assistance a customer might require at any specific moment. Considering the growing maturity of AI and analytics solutions today, it is not only possible to offer personalized solutions to each customer but also to do it at scale. It’s never a bad time to think out of the bot (pun intended). 

Where do ethics and accountability fit in?

If things go awry, it may be tempting to blame it on the algorithm, but that excuse is certainly not something we should be counting on. This is where the concept of algorithmic accountability comes into play. In a nutshell, it articulates the idea that the operators of the algorithm should appropriately set up sufficient controls to make sure that the algorithm performs as expected in the first place. 

Speaking of weaknesses that accompany AI, bias has also been a matter of concern in public discourse about implementing AI-based processes. Within the financial sector, particularly in assessing credit risk for instance, insufficient, unverified, or low-quality data could increase the likelihood of algorithms producing unfavorable conclusions for members of different religions, ethnicities, or to those belonging to minorities. Equipping AI algorithms with rich, verified, unskewed contextual data is one way of tackling this. The emerging field of Explainable AI or XAI is another promising development that can help us understand why and how a certain decision was made by the machine. The best part is that XAI makes it understandable not only for the experts but also for the common user, thereby helping institutions maintain transparency with their customers. 

Not without its challenges, but exciting times are ahead

We should however also keep in mind that developing and implementing AI-powered solutions in financial institutions, like in any other industry, is not without its challenges. Topping the list would be the challenge of recruiting and retaining AI experts, followed by inadequacies in technology to enable AI innovation and insufficient data sizes for model training and accuracy. That’s not to say that it’s without solutions. For instance, the growing area of industry-specific clouds and similar specialized services can help automate certain processes, thereby reducing the need for manual intervention. As these services are mostly designed to be easily deployed and customized, even organizations with limited AI expertise can leverage these services without needing to hire additional staff for these tasks. Development and data science resources thus freed up, will be able to focus on more strategic and differentiated areas. 

As we look around, we have so far reached a point in the evolution of AI that may have seemed impossible many years ago. Looking forward, however, nothing really seems impossible- the possibilities are limitless and the pace of development is unprecedented. It’s truly an exciting time for FinTech, to be able to extend the benefits of AI to the general public through ingenious interventions touching upon every aspect ranging from customer service and payments to lending and wealth management. Ultimately, the key is to strike a rewarding balance between the technological prowess offered by AI and the authenticity of human connection. 

About Autor

Anastasio Scalisi is the CTO of Revolv3. He brings over 25 years of international experience in the tech space. He has led technology transformation programs for enterprise media, telco and network companies. He worked as the interim CTO/advisor for Ingram Micro, leading major technology programs. He now runs his own technology advisory firm in Southern California. He enjoys skiing and snowboarding.