Google, IBM & Meta Certificates – 40% Off
One plan covers every Professional Certificate on Coursera.
Unlock All Certificates
Explore a 40-minute Black Hat conference talk that examines the security vulnerabilities in function calling (Plan & Execute) applications of generative AI. Learn about "PromptWare," a new class of zero-click input prompts that can manipulate GenAI engines to attack rather than serve applications. The presentation first covers basic properties of PromptWares before demonstrating how attackers can exploit known GenAI interfaces to trigger DoS attacks through infinite loops. Then discover Advanced Promptware Threats (APwT), which require no prior knowledge of the target application and leverage AI capabilities for reconnaissance, threat reasoning, and malicious execution—illustrated through an e-commerce chatbot example where attackers obtain unauthorized discounts. Presented by Ben Nassi (Infosec Researcher, Technion), Stav Cohen (PhD Student, Technion), and Ron Bitton (Principal AI/ML Security Researcher, Intuit).
