Four Ways That Generative AI Improved Cybersecurity Forever

InnovationFour Ways That Generative AI Improved Cybersecurity ForeverBySrinivas Shekar,Forbes Councils Member.for Forbes Technology CouncilCOUNCIL POSTExpertise from Forbes Councils members, operated under license. Opinions expressed are those of the author. | Membership (fee-based)May 18, 2026, 08:30am EDTSrinivas Shekar, Founder and CEO, Pantherun Technologies. getty​Use of AI in cybersecurity is not new. For a decade, Darktrace used unsupervised machine learning to identify network anomalies, for example. CrowdStrike ML analyzed behavior to classify and identify potential malware. And Google’s phishing detection AI became an industry benchmark.But traditional cyber AI relied upon deterministic rules. It detected and scored suspicious anomalies, but a human was still required to decide if action was required. Naturally, this produced an overwhelming burden of false positives for security teams to contend with. ​Generative AI drove a sea change in the security paradigm by performing synthesis and reasoning on its own, in natural language. Now, defenders can identify, interpret and respond to threats at machine speed. ​OK, so what does that really mean for the cybersecurity game? Here are four ways that generative AI changes cybersecurity forever:1. From Detection To Prediction​Traditional AI flagged anomalies only after something unusual occurred, generating alerts for humans to interpret and act upon, if required. That was quite useful, but at best, defenders could identify and thwart so-called “zero-day” attacks.Unlike older AI systems, generative AI can compare signals from across logs, failed access attempts and other reconnaissance activity to historical attack patterns, and then infer attacker intent with an accuracy as high as 85%. That alone represents a huge leap in defensive capabilities. However, the real linchpin here is that the AI can also predict the next steps the attacker is likely to take based on analysis far too sublime for any human to do in re...