AI-driven botnet detection in IoT networks: A comprehensive research review

Document Type

Article

Source of Publication

Computer Science Review

Publication Date

3-5-2026

Abstract

The intensifying expansion of Internet of Things (IoT) technology, projected to exceed 30 billion connected devices by 2030, introduces transformative benefits alongside significant security vulnerabilities. A critical concern is the pervasive threat of IoT botnets, which exploit the often-inadequate security postures of these devices to launch large-scale cyberattacks, including devastating Distributed Denial of Service (DDoS) campaigns. Despite the escalating danger, a systematic and comprehensive analysis specifically detailing the application of Artificial Intelligence (AI) for IoT botnet detection has been notably absent from the literature. This comprehensive review addresses this gap by classifying, critically evaluating, and synthesizing experimental research on AI-based methodologies for IoT botnet detection. Our study systematically investigates five core research questions: (1) the distinct phases of IoT botnet attacks, (2) conventional botnet detection methods and their limitations in the IoT context, (3) the spectrum of AI-driven approaches employed for IoT botnet detection, (4) a comparative analysis of these detection methods based on the datasets utilized, and (5) the evolving malicious activity scenarios within the IoT landscape. This study establishes a foundational understanding of AIpowered IoT botnet detection. Crucially, it identifies significant research gaps, including the persistent scarcity of diverse, real-world IoT botnet datasets, the imperative for more explainable AI (XAI) models in critical security applications, and the underexplored potential of federated learning and generative AI in creating resilient and privacy-preserving detection systems. These insights pave the way for future research aimed at fortifying IoT ecosystems against this evolving cyber threat.

ISSN

1574-0137

Publisher

Elsevier BV

Volume

61

Disciplines

Computer Sciences

Keywords

Internet of Things, Botnet, DDoS, Artificial intelligence, Security, Trust, Machine learning, Deep learning, Generative AI

Indexed in Scopus

no

Open Access

no

Link to Full Text

Share

COinS