AI-driven cyber threats and information sabotage: navigating data security and regulatory challenges

Abstract

The swift progress of artificial intelligence is fundamentally reshaping the cybersecurity environment, acting as a two-sided sword that empowers both defenders and adversaries. In 2025–2026, the United States alone reported over 3,300 data breaches, affecting hundreds of millions of individuals and resulting in considerable financial losses. Notable incidents, such as the PowerSchool breach affecting roughly 62 million students and teachers, AT&T data exposures, and major telecom incidents in Europe (e.g., Free in France and Dido in the Netherlands), highlight the growing risks to personal data security. This presentation studies the intersection of AI in cybersecurity, cyber management, and information sabotage. While AI strengthens defensive capabilities — including automated hazard detection and vulnerability remediation — it similarly amplifies offensive tactics such as hyper-personalized phishing, adaptive malware, deepfake-driven social engineering, and automated exploit chaining. At the same time, information sabotage and fake news have become powerful hybrid threats, often spread through botnets, deepfakes, and compromised media channels, weakening public trust, influencing elections, and enabling corporate espionage or financial fraud. From a cyber management perspective, organizations need integrated strategies covering risk assessment, incident response, and supply chain security. Regulatory compliance makes the picture more complex, with frameworks such as the EU’s NIS2 Directive (including its 2026 amendments), GDPR, and the Cyber Resilience Act requiring timely incident reporting, strong risk management, and certification. Failure to comply can lead to heavy fines and brand damage, whereas effective adherence offers a competitive edge.