Exploiting AI and Machine Learning Systems: Adversarial Attacks against Deep Learning Algorithms

No ratings

Presented at CactusCon 13 by

Thomas Wilhelm

The exploitation of vulnerabilities in AI and machine learning systems has significantly hindered the development of AI-based diagnostic and detection solutions across various industries - including automotive, healthcare, cyber security, and physical security. In this talk, we will discuss these critical vulnerabilities and present real-world case studies that highlight the challenges and life-threatening risks faced by researchers and developers. Furthermore, we will offer practical advice on how to secure AI systems for those organizations and individuals interested in developing safe machine learning solutions.