AI Security

Agentic Cyberattacks: The Next Evolution of Cyber Threats Introduction For years, security teams have focused on AI-assisted cyberattacks—situations where attackers use AI to write better phishing emails, generate malicious code, or automate repetitive tasks. That era is already behind us. We are now entering the age of agentic cyberattacks, where AI doesn’t just assist attackers. It acts. Agentic cyberattacks involve autonomous AI agents capable of planning, executing, and adjusting multi-step intrusions without waiting for human direction. These attacks are strategic, adaptive, and continuous, and they represent one of the most significant shifts in cybersecurity we’ve seen in decades. ...

ETR-6: Agentic Cyberattack Simulation Lab Overview This lab introduces students to the concept of agentic cyberattacks by simulating a full intrusion chain using a reasoning-only AI adversary. The AI model never executes real attacks. Instead, it explains how an autonomous adversary would think, plan, adjust to defenses, and optimize its operations during a fictional intrusion. Students will guide the simulated adversary through seven phases: reconnaissance, vulnerability reasoning, adaptation, phishing, persistence, lateral movement, and attack chain optimization. At every phase, students will analyze the agent’s decisions and map defensive strategies that could disrupt or detect the intrusion. ...

Securing the Future: A First Look at SANS Critical AI Security Controls v1.1 Introduction Artificial intelligence is no longer experimental—it’s embedded in production workflows, critical infrastructure, and high-stakes decision-making. But while AI evolves rapidly, security practices have struggled to keep pace. Enter the SANS Critical AI Security Controls (v1.1)—a new framework aiming to establish foundational, adaptable, and actionable controls for securing AI systems. This blog breaks down the key components of the SANS draft, explains their relevance, and suggests first steps organizations can take today. ...

AI’s Dual Role in Cybersecurity: Defense Tool or Attack Vector? Introduction Artificial Intelligence (AI) is no longer a futuristic concept—it’s a core component of modern cybersecurity. On one hand, AI empowers defenders to automate threat detection, identify anomalies, and respond to incidents faster than ever. On the other hand, attackers are weaponizing AI to scale attacks, bypass defenses, and even generate malicious code in seconds. This post explores both sides of AI in cybersecurity, provides practical defensive strategies, and introduces a mini hands-on lesson to get you started building more secure AI-integrated environments. ...