GENERATIVE AI FOR INTELLIGENT THREAT DETECTION: A HYBRID DEEP LEARNING FRAMEWORK FOR SECURE DIGITAL SYSTEMS

Abstract

As organizations undergo quick digital transformation and cyber threats grow in sophistication, there is a growing need for intelligent, adaptive, and trustworthy cybersecurity solutions. Traditional threat detection technologies are prone to failing to detect new attack vectors, zero-day exploits, and sophisticated cyber attacks because they are based on signatures and only adapt to known threats. The recent developments in Generative Artificial Intelligence (Generative AI) and deep learning offer new avenues to improve intelligent threat detection, including dynamic pattern recognition, anomaly detection, and predictive cyber defense. Prior work tends to emphasize the detection performance of AI-based cybersecurity models while largely neglecting the strategic use of Generative AI in more secure and explainable threat detection. This study aims to fill this gap by conducting a qualitative Systematic Literature Review (SLR) on the use of Generative AI for creating intelligent and secure digital systems. The research design is qualitative with an interpretivist paradigm, and the study is systematic in synthesizing the major peer-reviewed literature from the various academic research databases to gain insight into trends, methodological advances, challenges, and future research directions. The results highlight five predominant themes: the disruptive impact of Generative AI on cybersecurity, hybrid deep learning solutions for intelligent threat detection, explainability and trust in AI-powered security systems, hurdles in secure digital environments, and future prospects for adaptive cyber defense. The review shows the benefits of combining Generative AI and hybrid deep learning methods in boosting detection accuracy, bolstering resistance to advanced cyber threats and aiding in proactive security decision-making. The study extends the ongoing debate on intelligence in cybersecurity in both theoretical and practical aspects, and offers insights for researchers, practitioners, and policymakers to design intelligent security systems that are transparent, adaptive, and trustworthy. This paper stands out for its qualitative synthesis of existing evidence and the suggestion of a conceptual framework to support the integration of Generative AI into hybrid deep learning systems, thereby enhancing secure digital systems. The study overall finds that Generative AI holds significant promise for changing the paradigm of intelligent threat detection, and for fostering resilient, secure, and trustworthy digital ecosystems.