Voice Phishing Prevented by 'Technology' Not 'Wits'... Banking Sector Takes Action

[Asia Economy Reporter Yu Je-hoon] To prevent the increasingly sophisticated telephone financial fraud (voice phishing) damage, the financial sector is deploying various new technologies such as artificial intelligence (AI). Since there are limits to preventing damage through individual ‘wit (機智),’ the aim is to block it at the system level.

According to the financial sector on the 8th, Shinhan Bank introduced the banking industry's first ‘AI Abnormal Behavior Detection Automated Teller Machine (ATM).’ The AI abnormal behavior detection ATM detects when a customer exhibits abnormal behavior during an ATM transaction, such as talking on a mobile phone or wearing sunglasses or a hat, and provides a cautionary message to the customer before the transaction.

To this end, Shinhan Bank underwent about eight months of AI deep learning with an external AI company to learn and analyze transaction types by age group. Based on this, the bank decided to introduce AI abnormal behavior detection ATMs at branches with many elderly customers and a high risk of voice phishing incidents. Shinhan Bank plans to expand this to all ATMs in the future.

Additionally, in the second half of the year, Shinhan Bank plans to analyze the correlation between abnormal behavior detection data and accounts involved in voice phishing incidents, linking it with the Fraud Detection System (FDS). If AI detects abnormal behavior, the process will be changed so that additional identity verification is required before the transaction can proceed.

Shinhan Bank is not the only one rushing to develop technology to prevent voice phishing incidents. Industrial Bank of Korea (IBK) also built and piloted a ‘Telecommunication Financial Fraud AI Monitoring System’ using AI. This system prevents damage by automatically stopping payments when a customer suspected of voice phishing, classified by behavior and transaction patterns, attempts to withdraw more than 5 million KRW in cash at a branch. During the three-month pilot operation, it prevented a total of 750 cases, amounting to about 6.2 billion KRW in financial fraud damage.

NH Nonghyup Card recently established a voice phishing detection system linked with NH Nonghyup Bank accounts. When an abnormal transaction suspected of voice phishing is detected for NH Nonghyup Card users, the system additionally verifies the customer’s account information and takes action. It also includes a function that, during the long- and short-term card loan screening process, if suspicious patterns indicating fraud are detected, the loan is either rejected or the loan amount is transferred with a delay.

As the financial sector’s efforts to prevent voice phishing continue, related damages are also decreasing. According to the Financial Supervisory Service, as of 2020, the number of voice phishing incidents and the amount of damage were 25,859 cases and 235.3 billion KRW, respectively, down 64.3% and 65.0% from the previous year.

This content was produced with the assistance of AI translation services.