AI Deception: Fooling (Artificial) Intelligence Is Easier than You Think

Posted on in Presentations

Dolphin attack and kangaroos that confuse self-driving cars present cyberthreats on AI systems. Like other notable technologies, AI systems are getting security traction and are found to leave undefined edge cases—the material from which cyberattacks are made. Based on experience, this session will provide effective tips for secure incorporation of learning technology in adversarial environments.

Learning Objectives:
1: Learn the cyber-lifecycle and use wider perspective on security and technology.
2: Assessment and mitigation of the risk that comes with AI technology.
3: Better understanding of secure incorporation of AI technology.
Itsik Mantin


Lead Scientist, Imperva

Share With Your Community