Agentic AI Malware - Why the Cybersecurity Battle Isn't Over

Explore the emerging threat of autonomous AI-powered malware in this 46-minute conference talk from BSidesLV. Examine how agentic AI systems can be weaponized for cybersecurity attacks through detailed analysis of autonomous malware capabilities. Witness a proof-of-concept demonstration featuring a PowerShell agent equipped with metamorphic obfuscation techniques that can dynamically alter its code structure to evade detection. Learn about reasoning models that enable malware to develop and adapt attack strategies independently, making traditional defense mechanisms less effective. Discover comprehensive mitigation strategies designed to counter these advanced AI-driven threats, including detection methods and defensive frameworks. Gain insights into the current limitations of agentic AI malware and understand why, despite its sophisticated capabilities, it has not yet become a decisive factor in the cybersecurity landscape. Understand the implications of AI-powered autonomous agents in malware development and their potential impact on future cybersecurity challenges.