Artificial Intelligence (AI) and Machine Learning (ML) technologies continue to expand in their applications, uses, and benefits for credit unions. Because of this maturity and expanded adoption rate, AI/ML is helping to solve highly complex opportunities that generate positive ROI across business segments.
Credit unions are increasingly interested in deploying these technologies across their businesses to support areas such as risk management, reducing friction in loan origination departments, income and verification controls, fraud reduction, and compliance and auditing processes.
Ultimately, credit unions continue to strive toward lowering the cost of credit using AI/ML for real-time transparency, greater financial inclusivity, and improved compliance. Here are some critical use cases of how financial institutions are leveraging AI/ML in 2023.
Conversational chatbots
Conversational chatbots help lenders interact with customers in a more conversational way. Consumers desire the same level of customer service they receive from leading tech-forward companies like Amazon, Netflix, and Lyft.
AI-driven chatbots and virtual assistants offer 24/7 assistance to customers on many items such as account balances and recent transactions. What’s most impressive is that these chatbots enable customers to send funds using conversational language.
Customer sentiment analysis
For many years financial institutions had a difficult time combining customer sentiment into their big data and automation platforms. Today’s leading lenders have access to a plethora of data about their customers, but historically a large portion has been unstructured and difficult for computers to understand.
AI, however, can analyze what customers communicate and pinpoint the emotions they are expressing in real-time. These systems can alert lender customer service teams so that they can resolve issues effectively and faster.
Creditworthiness for a thin file or no file
AI/ML also help provide a clearer view of a customer’s creditworthiness, especially when they have a thin file of credit, no file of credit, or if they have supplemental sources of income, such as many of today’s gig economy workers.
Let’s take a closer look at a specific use case of AI/ML in credit union automotive finance since credit unions have experienced growth in that area. During Q4 of 2022, credit unions originated 27% of all auto loans and leases, making them the largest share of the market. Additionally, they financed more used vehicles than any other type of lender, accounting for 31% of used vehicle financing during the fourth quarter.
How AI identifies loan defects in automotive finance
The Consumer Financial Protection Bureau (CFPB) has increased its level of scrutiny on the accuracy of loans and the paperwork documentation (called deal jackets) that takes place between a credit union and a dealership. In many cases, audits take place to investigate if a lender may have misrepresented costs in loan agreements that may have placed customers in high-cost loans for cars in violation of the Consumer Financial Protection Act of 2010.
The scenario represents one of the latest examples of regulators pushing the boundaries by introducing new laws or enforcing existing ones which leverage interpretations that place administrative pressure on credit unions and their compliance teams. Many credit unions remain susceptible to fines and penalties that are detrimental to their operations and bottom lines.
Credit unions can more stringently mitigate these scenarios through the implementation of AI-powered systemic controls that help them avoid this additional scrutiny and audit environment. Today’s AI-powered software enables credit unions to comply with regulatory requirements and be audit-ready. The solutions offer policies that are clear and standardized, and lenders are guided through model governance compliance for internal audits while providing expert advice and sample documentation, if necessary.
Using AI model documentation
Model documentation from today’s AI software includes a qualitative assessment of the potential for disparate impact risk in the models built for credit unions. The auditing process performs quarterly, quantitative disparate impact assessments. The analyses are based on race, ethnicity, gender, and age (62+), and while the process doesn’t collect race and ethnicity data, it does employ the CFPB’s Bayesian Improved Surname Geocoding (BISG) proxy method for race, ethnicity, and gender using the most recent census data.
The software today leverages advanced AI technology to simplify and automate the process of collecting and analyzing data, with the goal of helping to fund loans as quickly and efficiently as possible while lowering the cost to fund, lowering the cost of processing GAP refunds for early payoffs, improving compliance, and lowering the cost of regulatory Matters Requiring Attention (MRAs) and consent decrees related to unfair, deceptive, or abusive acts and practices (UDAAPs).
Like financial providers across all industries, credit unions are not AI/ML experts, and it’s not their core competency, so they understand the importance of finding quality outside experts in AI/ML today who can help. Trusted partners are being tapped to help catch these loan defects, where improper deals can be flagged that are not ready for funding.
AI software allows funders to focus on complete deals, enabling their teams to quickly address any identified issues with dealers. It also enables automation of dealer defects, instantly notifying dealers of document defects to reduce contracts-in-transit, fund deals faster, and reduce compliance and regulatory risk.
It is also important to note that AI and automation are increasingly being deployed for auto lenders and credit unions outside of simple loan defects. A recent survey of lender executives found that 63% plan to implement AI and automation technologies this year for securitization, 61% for loan servicing, and 52% for loan processing and finding.
Implementation is growing
While AI and ML are still in their infancy stages for financial services providers, the adoption of these technologies continues to grow. More importantly, these institutions are realizing the positive impact it has on their operational bottom line, employee morale, and overall customer experience.
