ÁùÉ«Íø

ChatGPT Unchained: The Art of Hacking Models

ChatGPT Unchained takes participants on a thrilling journey into the vulnerabilities of artificial intelligence. Attendees will learn how to craft adversarial examples that demonstrate the fragility of AI's perception. By manipulating input data, participants will observe real-time how slight changes can dramatically alter model outputs. The session includes a practical exploration of model poisoning, where subtle corruptions during training lead to malicious outcomes. Finally, participants will engage in a guided exercise to implement defense mechanisms, highlighting the importance of robust AI security. 

Session Presenters

David Nimri

CIS Instructor

The Chaffey Cybersecurity Center supports the Chaffey Cybersecurity Club which allows them to participate in cybersecurity competitions and host events. Chaffey College Cybersecurity Club is learning and growing in the cybersecurity field.