Exacerbation and Combat of Cyberattacks
Devika Sharma, Saket Sharma · 2025
Machine learning is a subset of artificial intelligence. It is one of the most renowned present technological manifestations of the fourth industrial revolution or Industry 4.0. It permits systems to get trained from past data and improve. Machine learning is used to analyze cybersecurity data. Ironically, it is the same feature of artificial intelligence that cybercriminals exploit for their malicious ends and security professionals use for predicting cybersecurity solutions. This reflects the dual paradox of artificial intelligence. Artificial intelligence is predicated on machine learning models and has the potential to identify trends via algorithms and forecast future actions. This enables the preemption of ill-intentioned behavior. One of the most common misuses of AI is “deep fakes,” which are based on the use of AI to distort audio and visual data, such that they appear authentic. Deep fakes spread misinformation and are difficult to distinguish from genuine content. AI-powered password guessing is another threat area, where cybercriminals use machine learning to develop more sophisticated algorithms to guess user passwords. In addition to traditional approaches, “generative adversarial networks” (GANs) equip cybercriminals to examine large password datasets and engender password variations that align with the statistical distribution. Another form of the misuse of AI is to utilize it to impersonate human behavior on social media. Cybercriminals can cheat systems of bot detection on social media platforms, like Spotify, by imitating “human-like” usage trends. Thus, cybercriminals can monetize the illegitimate system to create fraudulent streams and traffic for an artist. Similarly, cybercriminals are exploiting AI to attack gullible hosts. Thus, phishing, business email compromise scams, and manipulation of cryptocurrency trading practices are some manifestations of AI abuse by cybercriminals. They rely on AI to upscale the magnitude and ambit of their attacks. In keeping with the latent potential of AI and its paradoxical role, it can identify patterns of cyberattacks early and contribute to cybersecurity. It can furnish information about cyber threats, which can provide insights into how these professionals prioritize digital assets more likely to get violated. Thus, AI is integral to “vulnerability scanning and pentesting.” A technique of machine learning called “user and event behavior analytics” (UEBA) spots abnormal behavior and highlights it as suspicious. Machine learning can be used to diagnose phishing attempts and similar attacks by tracking threats in real time. Cybersecurity and data governance are significant at present. AI and machine learning are becoming indispensable in cybersecurity, in a scenario where security experts face manpower crunch and time constraints. In the proposed chapter, we will be analyzing the current cyber threats, the role of AI in aggravating them, and how AI can address these threats and ensure cybersecurity. We will also examine the legal and policy responses to cyberattacks in the Indian context.