Auto lenders are incorporating artificial intelligence into their processes to improve customer service automation and credit decisioning while eyeing uses for underwriting.  

Subprime auto lenders can use AI to ensure staff and resources are assigned to tasks that help navigate affordability challenges, operational costs and credit risk, Harvey Singh, chief operating officer at Veros Credit, said during a webinar on the impact of conversational AI hosted by fintech Skit.ai on Sept. 19.  

“We have taken the approach of making sure our core staff is handling the nitty-gritty and the harder aspects of the business,” he said. “AI [is] handling the low hanging fruit. … We’re not bombarding our resources with the repetitive tasks.” 

Santa Ana, Calif.-based subprime auto lender Veros Credit has ramped up use of generative AI in the past six to eight months, largely for customer service automation, Singh said. Veros has implemented AI for consumer communication tasks such as assisting with payoffs and providing information quickly, he said.  

“We can service our customers 24/7, on the weekends and after hours when we are not available,” Singh said. “Customers are struggling, and delinquency is rising, and our goal is, how can we service this customer and stay connected when we are not here? Gen AI is helping us to communicate and pick up that information and respond back accordingly.” 

The caveat is that Veros must consistently monitor AI-based communications to ensure the bot is “behaving the way it should be” and providing accurate information, Singh said.  

“I don’t think AI will replace humans, but it’s a supplement to enhance your ability to handle more while you continue to grow,” he said, adding that hesitation to adopt AI in the industry is largely due to costs and staffing concerns.  

AI as a tool 

Generative AI must be seen as a tool to supplement customer service, Armando Hidalgo, director of servicing at Irving, Texas-based subprime lender SameDay Auto Finance, said during the webinar.  

“We deployed [AI] in our servicing department, which also takes care of customer service and the rest of the portfolio, as a helping hand for our staff,” he said. 

Hidalgo said lenders should consider these questions when incorporating AI: 

• What is your desired result? 

• Are you looking to make more contact?  

• Are you looking to scale up or scale down? 

AI in underwriting 

AI can also assist in credit decisioning by identifying hidden factors such as buy now, pay later debt or income from non-traditional work such as content creation on social media, Hidalgo said.  

“If you have machine learning and AI … you are able to make a better [credit] decision,” he said.  

Veros Credit is looking at ways to use AI in underwriting to fund deals faster and “remove some of the repetitive tasks and give it to AI so our core staff can focus on the tasks that are necessary and required,” Singh said.  

“In the next six to 12 months, we’d like to double and triple the areas where we are using [AI],” he said, noting that remaining compliant while building out AI uses is critical. 

“It was challenging for us to adopt AI. We are conservative, but we were able to put controls in place,” Singh said. “Our main focus [is] making sure that compliance comes first before we start making adjustments in any other areas.” 

As an emerging technology, lenders must be mindful of how it is integrated into existing processes, Hidalgo said. 

“This new era of AI — whether the benefits are substantial — [there is a] learning curve,” he said. “It can be expensive, so implementing proper installation is key. You have to keep your ear to the ground, your finger on the pulse and continually review these things.”  

Editor’s note: This article first appeared on Auto Finance News, a sister publication to Bank Automation News.