"ORNL Malware 'Vaccine' Generator Licensed for Evasive.ai Platform"
The use of Artificial Intelligence (AI) and Machine Learning (ML) is transforming technology and product development, resulting in the creation of more intelligent, efficient, and tailored applications that use large amounts of data. However, the same capabilities also benefit malicious actors, who use AI to build malware that evades detection by the algorithms employed by network security solutions. A technology created at the Department of Energy's Oak Ridge National Laboratory (ORNL) and used by the US Naval Information Warfare Systems Command (NAVWAR) to test the capabilities of commercial security tools has been licensed to the cybersecurity company Penguin Mustache for use in the development of its Evasive.ai platform. Creator of the technology and former ORNL scientist Jared M. Smith launched the company alongside his business partner Brandon Bruce. Smith, who worked for six years in ORNL's Cyber Resilience and Intelligence Division, developed the Adversarial Malware Input Generator (AMIGO) at the Department of Defense's request. AMIGO was developed as the evaluation tool for a challenge set by NAVWAR involving AI applications that autonomously detect and quarantine cyber threats. AMIGO generates optimally evasive malware alongside the training information required for a security system to detect it in the future. It uses over 35 million malware samples, some of which are publicly available and others never before seen. This article continues to discuss AMIGO and its licensing to Penguin Mustache.