AI Didn’t Break Cybersecurity — It Changed the Rules

AI is no longer just a tool — it’s both a defender and an attacker. And cybersecurity today looks less like a firewall problem and more like a chess game. In this presentation, I look at cybersecurity through managerial and AI optics, not from a purely technical or SOC-level perspective. Every move creates a counter-move: Attackers use AI for deepfakes, social engineering, adaptive malware, and automated reconnaissance Defenders respond with AI-driven detection, automation, and decision support Governance, judgment, and timing increasingly matter as much as technology This session explores: Why AI is simultaneously a friend and a foe in cybersecurity How large language models introduce new attack vectors such as prompt injection and agent abuse Why AI risk is now overtaking ransomware as a top concern for security leaders The paradox of rapid AI adoption without mature governance How frameworks like NIST AI RMF, ISO/IEC 42001, and the EU AI Act change the role of managers Why cybersecurity decisions are increasingly management decisions, not just technical ones