What FIs Get (And Don’t Get) About AI

The jury is in. Almost every modern financial institution (FI) has at least one artificial intelligence (AI) or machine learning system in place, and 61 percent intend to invest in more. The only question is: What learning systems will they choose to adopt?

Among the many AI and machine learning systems available on the market, there is one particularly versatile type of AI, having been engineered specifically to enhance financial and payment operations: smart agents.

Smart agents can adapt to complex ecosystems, and be assigned to perform particular functions. One smart agent may be assigned to analyze the actions of a consumer, point-of-sale (POS) terminal or ATM, and use that analysis to learn how to best perform whatever action it was assigned to do.

This ability to analyze and learn from an experience makes smart agents highly effective tools in the financial services space. It’s no wonder that 64 percent of all FIs said they would be interested in using them.

So, in what areas are smart agents most effective, and how do these businesses hope to leverage the tech?

In the latest edition of the AI Innovation Playbook, PYMNTS teamed up with Brighterion to survey more than 200 American FI decision-makers and gauge what they stand to gain from incorporating smart agent technology into their everyday operations. According to the research, FIs interested in smart agents want to deploy them in almost every business area imaginable — and different types and sizes of FIs want to use them in varying degrees and ways.

For instance, while 72.1 percent of commercial banks are interested in using smart agents to enhance their banking services, just 53 percent of credit unions are interested in doing the same. Meanwhile, 100 percent of FIs with more than $100 billion in assets are interested in adopting smart agents to support credit underwriting, and 89.5 percent of those with assets valued between $25 billion and $100 billion are interested in them to help in the fight against internal fraud.

Despite smart agents’ many use cases, 36 percent of FIs are either only slightly or not at all interested in smart agents. Their reasons often have less to do with the technology itself, and more to do with their own limitations. For example, 50 percent of these FIs said their organizations lack the skill sets needed to handle them, indicating a staffing issue more than a technological one.

To understand what smart agents and AI systems can offer, and what other learning systems can’t, download the Playbook.