The Financial Services Commission held a meeting with related authorities, private sector experts, and financial industry associations on July 28 and introduced measures to step up anti-vishing response efforts and set up a comprehensive anti-vishing platform using artificial intelligence (AI) analytics.

Anti-vishing AI Platform

Currently, individual financial companies are able to take appropriate anti-vishing measures on suspicious account activities based on the operation of their own fraud detection systems (FDS). However, the current response system faces hurdles in its capacity in that individual financial companies have experience with only limited numbers of prior vishing cases and carry out surveillance only based on their own analysis of scam patterns. Moreover, even when they detect an account used for vishing scam, there is no instant information sharing between financial companies, so that they lack sufficient information about the newly emerging types and patterns in vishing scams individually. This has also created significant divergence between financial companies in their anti-vishing response capacities and has not helped to speed up the process of suspending problematic accounts used in criminal activities.

In this regard, the anti-vishing AI platform is a digital data infrastructure that will help the authorities to tackle these limitations in the current anti-vishing response system. The platform will gather information and data on suspicious accounts from across all financial sectors, telecom service providers, and investigative authorities and will operate two different types of information categories—those required for immediate sharing and those analyzed by AI.

The types of information categorized as requiring immediate sharing will be instantly disseminated and shared with financial companies and related organizations to facilitate immediate suspension of account activities. This will help to quickly freeze multiple financial accounts used in criminal activities across different financial sectors at once. It will also help to prevent further damages for victims and swiftly block escape routes for criminal proceeds.

The types of information categorized as AI analytics collect and provide information focused on specific characteristics of suspicious accounts used in vishing scams from financial companies. After going through an AI-based pattern analysis on the latest vishing scams conducted by the Financial Security Institute (FSI), the information will then be used by all financial sectors to prevent and block criminal accounts in advance. Through the sharing of AI-analyzed information, financial companies, telecom service providers, and investigative authorities will be able to freeze account activities in advance on suspicious accounts, warn and notify customers about suspicious phone numbers, and strengthen preventive measures for those deemed to be particularly vulnerable to vishing scams, respectively.

The AI analytics will also help nonbank financial companies to step up their anti-vishing response capacities. For telecom service providers and investigative authorities, the use of AI analytics will help to introduce new vishing prevention service functions and allow them to develop more targeted investigation strategies, respectively.

With the anti-vishing AI platform in place, financial companies, telecom service providers, and investigative authorities will be able to collaborate and coordinate and share relevant information on a more standardized and digitized manner, facilitating them to work together closer for speedier responses.

The anti-vishing AI platform is expected to be operational by the end of this year.

Further Plan

Alongside the launching of an anti-vishing AI platform, the financial authorities will seek to overhaul the current anti-vishing response system throughout the entire process—from preventing vishing scams in advance to blocking scams on the spot to providing remedies for victims to boosting advertisement efforts for raising public awareness.

* Please refer to the attached PDF for details.