As artificial intelligence becomes increasingly capable of understanding and generating computer code, an important question is emerging: can these same technologies help cybersecurity professionals identify malicious software and protect the digital systems people rely on every day? This question sits at the heart of Champie Joyce Maptue Tagne’s research into the intersection of artificial intelligence, software security, and malware detection.
In her peer-reviewed research published in IEEE Access, “Large Language Models for Malware Detection: A Systematic Review, Taxonomy, and Open Challenges,” Champie examined the rapidly developing use of Large Language Models (LLMs) in malware detection. Rather than focusing on a single model or experiment, her work brings together existing research to understand how LLMs are being applied to malware analysis, how approaches in the field can be systematically organized, and where important challenges remain.
The research comes at a critical moment. Malware continues to evolve while software systems become increasingly interconnected and complex. At the same time, LLMs are demonstrating new abilities to analyze code and reason about software behavior. Understanding where these models can contribute to malware detection and where their limitations may create risks is therefore becoming an important cybersecurity research problem.
For Champie, the study represents more than an examination of a new technology. It reflects a broader interest in understanding how artificial intelligence can be used responsibly in security-sensitive environments. Cybersecurity decisions require more than confident answers from an AI system; they require methods that researchers and security professionals can evaluate, understand, and ultimately trust.
One of the most important contributions of this research is the foundation it provides for future work. By systematically organizing the growing body of research on LLM-based malware detection, the study gives researchers a clearer picture of what has already been explored, the approaches being used, and the questions that remain unanswered. Its taxonomy can help researchers navigate an increasingly complex field, while the identified open challenges can point toward areas where further investigation and new solutions are needed. In this way, the research does not simply summarize existing knowledge; it can help researchers determine where the field needs to go next, providing a reference point from which new studies, methods, and security technologies can be developed.
The potential value extends beyond academic research. As organizations increasingly explore artificial intelligence for cybersecurity and software analysis, understanding the capabilities and limitations of LLM-based malware detection has practical significance as well. Insights from this research can help inform the development of future security tools and support more thoughtful evaluation of AI-assisted approaches before they are relied upon in real-world environments. Advancements in this area could ultimately contribute to more effective malware analysis, stronger software protection, and better-informed cybersecurity practices across organizations that depend on increasingly complex digital systems.
The publication also forms part of Champie’s broader research journey at the intersection of cybersecurity, artificial intelligence, and software engineering. Her work has explored Android application security, software quality, malware detection, and the reliability of AI-assisted security analysis, areas connected by one central concern: how can increasingly intelligent software systems be used to make the digital world more secure without creating a false sense of security?
As artificial intelligence becomes more deeply embedded in cybersecurity, the importance of research that maps what is known, identifies what remains unresolved, and guides the development of future solutions will continue to grow. By providing a structured understanding of the emerging field of LLM-based malware detection and highlighting directions requiring further investigation, Champie’s research offers researchers and practitioners a foundation they can build upon as they work toward more capable, reliable, and trustworthy approaches to detecting malicious software. Its broader significance lies not only in understanding the technology available today but also in helping shape the research and innovation needed to build the cybersecurity defenses of tomorrow.
