Anti-Tamper in an AI World
Anti-Tamper in an AI World
Anti-Tamper in an AI World
In the not so distant future, maintaining superiority will require critical systems to extend far beyond sensor fusion and into real-time learned adaptation within a theater of operation. This requires secure system-of-systems architecture to dynamically learn from and collaborate with each other on-demand, all while self-modifying as needed.
Secure AI Collaboration
Anti-Tamper in an AI World
Anti-Tamper in an AI World
Our new research area centers on addressing the threat space of two or more independently designed AI-based system-of-systems and how they can autonomously collaborate to create new machine behaviors. How can we robustly defend AI machines from AI-based attacks using a new generation of dynamically learning Anti-Tamper countermeasures?
