AI Biggest Payments Impact May Be Helping Humans Manage Risk

The financial services and payments sectors have long been leveraging automated tools to streamline operations.

Now, those legacy tools and intelligent solutions are getting exponentially smarter as advances in artificial intelligence (AI) allow all of the various technologies that fall under its broad moniker to do more with less.

“We really are surrounded by AI in our daily lives,” Michael Haney, head of Cyberbank Digital Core at leading FinTech platform Galileo, the sister company of Technisys, said during a recent conversation with PYMNTS. “In some ways it goes back to the emergence of business rules engines in the 1980s and the applicability of those in the banking industry, where they were often embedded into workflow and case management tools.”

Then, he added, those business rules engines suddenly got a lot more sophisticated.

“There’s an extension that some people lump under AI of a type of business process automation which we call robotic process automation that replaces or enhances what a human does by interacting with various systems and taking a lot of the low value tasks away for the humans so they can focus on more value-add tasks,” Haney said. “These robots, they can work 24/7 at a speed that you or I cannot work at.”

Leveraging Truly Intelligent AI

One of the bigger historical problems with rules engines that today’s AI tools have fixed, Haney said, is that machine learning solutions now have the capability to learn from and adapt to circumstances on their own by activating high dimensional data sets in what is called “deep learning,” rather than requiring a manual intervention when a process speed bump appears.

This ability of AI to move across multilayered data set “worlds” of images, speech, text and more is what makes today’s AI solutions worthy of being described as “intelligent.”

Talking about Galileo’s own conversational AI engine, Cyberbank Konecta, Haney said that it’s been a “fascinating evolution” from the early days of chatbots, which tended to follow more binary “press 1 or press 2” pathways.

“We have this deep learning technology called sentiment analysis that allows us to understand if the customer is getting frustrated,” Haney said. “It flags things like, are they getting angry? How are they feeling? And allows us to shift them to a human interaction that might be able to help the situation without leaving the channel or losing all the chat history.”

He noted that advances in AI along these lines are what is helping the technology evolve past automating human activity to “really augmenting human intelligence … allowing us to move beyond just things like operational costs and operational efficiency to more advanced areas of risk management, opportunity finding and forecasting, and regulatory compliance.”

Read more: How Generative AI Is Helping Generate New B2B Efficiencies

Getting Information to People Who Need It

The next big shift that Haney sees in how businesses can leverage AI solutions is what many in the industry call “next best action.”

“We’ve really started to see that pivot of how we think not just about uncovering patterns and trends but being able to take action on them in a very granular way, putting information into the hands of the people who need it most, when they need it most, so they can take the right action,” Haney said.

We can do even more intelligent things than just extract data, he added, saying that it’s now possible for AI solutions to understand context and take appropriate actions based on that context, which offers huge opportunities for customer personalization and tailored-fit product recommendations.

“We often only think of customer-facing versus staff-facing AI applications, but there will be lots of different flavors of AI helping out all kinds of divisions within the bank — as well as that bank’s end customers, as the technology continues to mature at a very rapid pace,” Haney said. “There’s applicability all over the place, it is not just all customer-facing.”

He notes that these are all very data-hungry situations. “Data is foundational to building the models, training the AI — the quality and integrity of that data is important, but even more important is the data privacy.”

Still, Haney said that while banks may struggle to train an AI to do something properly and compliantly, they also often struggle to train a human employee to perform the same task.

As for what he sees in the future?

“Hopefully the prices come down,” Haney said, noting that the smart way to take advantage of AI right now is to leverage experienced vendors in this space rather than attempting to build everything in-house.