Financial Security Institute's AI-Based Mobile App Security Risk Detection Technology Recognized at International Conference

Accepted as a Regular Paper at ISSTA 2026
17 Security Vulnerabilities Found in 50 Mobile Apps
Plans to Develop AI-Based Security Inspection Framework

Artificial intelligence (AI)-based technology for detecting security risks in mobile applications, developed in-house by the Financial Security Institute, has been recognized for its research achievements at a leading international academic conference in software testing and analysis. The technology is expected to be used in practical security inspections of mobile apps in the financial sector.

The Financial Security Institute said on Oct. 12 that a research paper on its AI-based technology for detecting security risks in mobile apps had been accepted as a regular paper at the international academic conference ISSTA 2026, where the research findings were presented.

The paper was co-authored by Song Heonghyeon, team leader in the AI Security Division at the Financial Security Institute, who served as first author. After a multistage review by experts from around the world, it was selected in June as a regular paper for the main research track of ISSTA 2026.

The research team presented its findings at ISSTA 2026, held in California, United States, from Oct. 4 to 9. The study drew particular attention as empirical research that identified multiple security vulnerabilities in real-world mobile apps.

The key achievement of the research was the development of technology that detects security risks by comprehensively analyzing not only the code inside mobile apps but also the behavior and information flows that occur during actual execution.

Mobile apps often use a variety of external libraries and operating system features, making it difficult to identify security risks during actual execution by examining only the code inside the app.

To address this, the research team supplemented the limitations of existing analysis methods by comprehensively tracking and analyzing behavior and information flows during app execution. Experiments on 50 real-world mobile apps identified a total of 17 security vulnerabilities, confirming the technology's effectiveness.

Based on these research findings, the Financial Security Institute plans to develop an "AI-based inspection framework" that can precisely analyze security risks in mobile apps and use it to inspect mobile apps in the financial sector. It also plans to continuously incorporate analysis results accumulated during inspections and new threat cases to improve the detection technology, while supporting efforts to raise the security level of mobile apps in the financial sector.

"These research findings demonstrate that the Financial Security Institute has the expertise to research security threats directly and develop its own technology to detect them," said Park Sangwon, president of the Financial Security Institute. "We will continue to turn outstanding research findings into security technologies that can be applied in real-world financial settings and provide practical support for improving the security levels of financial companies."

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