The Financial Security Agency has identified 17 security vulnerabilities in 50 mobile applications through the use of artificial intelligence (AI) technology. The agency plans to apply its proprietary detection technology to security assessments of mobile apps in the financial sector.
On October 12, the Financial Security Agency announced that a paper on its AI-based mobile app security risk detection technology was accepted as a regular paper at the prestigious software testing and analysis conference, ISSTA 2026.
The research was led by Song Seong-hyun, head of the agency's AI Security Division, who served as the first author. The paper underwent a multi-stage review by experts from around the world and was ultimately accepted for the main research track of ISSTA 2026 in June. The research team presented their findings at the conference held in California from October 4 to 9.
The core of the research involves tracking and analyzing not only the internal code of mobile apps but also the actions and information flow that occur during actual execution to identify security risks. The researchers noted that traditional code analysis has limitations in detecting vulnerabilities associated with external libraries or operating system functions.
By applying this technology to the 50 mobile apps, the research team successfully identified 17 security vulnerabilities, validating the effectiveness of their approach.
Based on these research results, the Financial Security Agency plans to develop an 'AI-based inspection framework' for security assessments of mobile apps in the financial sector. The agency aims to continuously enhance its detection technology by incorporating analysis results and new threat cases obtained during the inspection process.
Park Sang-won, the head of the Financial Security Agency, stated, "This research achievement demonstrates our capability to directly study security threats and develop our own detection technologies. We will advance these excellent research results into practical security technologies that can enhance the security levels of financial institutions."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.