gettyimagesbank

A growing number of commercial lenders are adopting artificial intelligence (AI) technology to screen high-risk corporate borrowers, cornered by criticism of imprudent lending practices despite steep loan defaults over the past few years, market watchers said Thursday.

Propelling the industry-wide move is a decrease in mortgage businesses, as mandated by tighter government lending regulations put in place to curb the country's world's-highest household loans.

Some say a data-oriented credit score assessment will bolster the financial soundness and structural efficiency of lenders, long gripped by lingering capital deterioration risks. Others say machine-mediated approvals will erode the prospects of small and medium-sized enterprises, exacerbating the already-prevalent business-of-scale disparities.

"Our machine learning-incorporated credit evaluation will process tasks previously undertaken by human beings — only faster," an NH NongHyup Bank official said.

This was in reference to the bank's recently introduced AI-operated corporate credit evaluation system.

Reviewed will be financial data, including revenue, operating income, net income, cash flow and collateral. The growth prospect of the industry and payment history will be considered as non-financial alternative data.

"We plan to expand the system to enhance overall productivity," the official said.

The bank dispelled criticism that the new AI system will prioritize quantitative data over qualitative data, a task long handled by human workers.

"The new system does not create hurdles whereby potential borrowers are filtered out and denied loans," an NH official said.

"Firms that meet the criteria will be granted loans quickly, unlike others that need on-site inspections by bank officials or internal discussion."

Similarly, KB Kookmin Bank utilizes the Big Data Credit Scoring System, an AI-driven service introduced in 2022.

It integrates financial and non-financial information for bank employees to evaluate and approve loans. The system can also provide a list of firms with high growth potential.

The AI system is updated regularly through self-learning to best reflect the changes in economic and financial conditions every year.

Shinhan Bank uses AI in loan evaluations and fraud detection services.

The AI-mediated lending program has its strengths, according to Suh Jeong-ho, head of the demographic change center at the Korea Institute of Finance.

"Naver introduced an AI loan evaluation model whereby potential borrowers — sellers on Naver's shopping platform — are granted about 20 million won ($14,000) in loans," he said.

Referenced for credit review were their social media presence, positive product reviews, number of refunds or exchange requests and returning customers.

"The new criteria by the AI program allowed many small budding merchants with short track records of business to take out small loans, long denied by commercial lenders," Suh said.

The outstanding balance of corporate loans extended by the county's top five banks — KB Kookmin, Shinhan, Hana, Woori and NH NongHyup — came to 829 trillion won as of Monday, up 8.6 percent from the previous year.