Agentic AI: The Weapon That No Longer Needs a Warrior
Agentic AI can carry out security tasks as well as suggest them. This essay examines what that shift means for attackers, penetration testers, and the people responsible for checking the results.
Articles, essays & research
Writing & Research
Essays and research on AI security, human judgment, and adversarial assessment.
Offensive AI: The Sine Qua Non of Cybersecurity
Foster argues that defenders need to study offensive AI to understand the attacks it enables. This contributed article connects earlier advances in malware and defensive tooling with the challenges posed by generative AI.
The Evolution of the Digital Predator: Using AI to Evade Security Controls
Foster’s 2023 SANS research examines whether an operator with limited technical skill can use generative AI to produce malware that evades traditional security controls, and what that capability means for defenders.