: This review charts the evolving intersection of artificial intelligence and cybersecurity, showing how machine and deep learning techniques are redefining intrusion detection and threat-hunting. Zhao et al. first outline the algorithms applied to spam, phishing, DGA, malware, and botnet detection, then highlight the obstacles that plague real world applications. To overcome these hurdles, the authors design a repeatable detection workflow that fuses multiple log sources and labels them. A feed-forward neural network trained on these automatically labeled windows is not only accurate, but linearly scalable. The paper argues that AI-driven alerting is only the first step-full incident response still demands automated correlation and analyst investigation.
Author:
Liguo Zhao, Derong Zhu, Wasswa Shafik , S Mojtaba Matinkhah, Zubair Ahmad, Lule Sharif, and Alisa Craig
