Banks ramp up AI defenses as gov’t moves to make them liable for voice phishing losses
Yoon Chang-ryul, minister for government policy coordination, along with other senior officials, announces comprehensive anti-voice phishing plans at Government Complex Seoul, Thursday. Yonhap
Banks are enhancing preventive measures against voice phishing, including upgrades to monitoring systems powered by artificial intelligence (AI), as the government moves to require financial institutions to compensate phishing victims, industry officials said Monday.
The efforts come as the government plans to legislate a no-fault liability rule, which would require financial institutions to compensate victims of voice phishing even when transfers were made directly to criminals by the victims and even if the banks are not at fault.
This measure is a core element of the comprehensive anti-voice phishing plans prepared by a pan-government task force led by the Office for Government Policy Coordination, with participation from the Ministry of Science and ICT, the Financial Services Commission (FSC) and other agencies.
In announcing the plans on Thursday, the government said it will set detailed requirements, limits and procedures in consultation with financial firms as early as this year.
Accordingly, the country’s major banks plan to focus their follow-up measures on strengthening defenses against voice phishing. They believe their compensation burden would increase considerably after damages occur, making preemptive fraud prevention crucial. To that end, they are placing greater emphasis on deploying AI technology to enhance protection.
KB Kookmin Bank operates both a fraudulent transaction detection system and an AI-powered voice phishing monitoring system, using an AI relearning model that autonomously gathers data and retrains itself to respond swiftly to new phishing tactics.
Shinhan Bank is rolling out dedicated desks for telecom financial scams across all 652 branches, where trained staff can immediately freeze fraudulent accounts when victims seek help. It has also partnered with crypto exchange Korbit to build a joint defense system, reflecting the rise in laundering of stolen funds through digital assets.
Hana Bank has developed a next-generation detection platform that combines historic fraud scenarios with AI deep learning, which the bank says has been highly effective in preventing both phishing and fake account fraud. Additionally, the bank’s app alerts users if a phishing app is detected on their devices at login.
Woori Bank and NH NongHyup Bank are also enhancing protections through collaborations and upgraded monitoring tools.
Woori has teamed up with the police and the Financial Security Institute to manage suspicious overseas accounts, and partnered with mobile carrier LG Uplus to share scam techniques, develop joint initiatives and formulate detailed action plans through a dedicated working group.
NongHyup, meanwhile, is refining its monitoring models based on the latest fraud cases and operates 24-hour surveillance on suspicious accounts.
“As voice phishing tactics grow more advanced, individuals alone cannot effectively prevent losses,” an FSC official said. “Financial institutions, with their specialized expertise and infrastructure such as fraud detection systems, must now share responsibility and implement a more coordinated response.”