Privacy-aware interactions with Generative AI

This thesis explores the complex dynamics of privacy arising from human interaction with multimodal Generative AI systems. The student will design and develop an innovative prototype of a system/tool that can analyze and flag potential privacy risks in real-time during interactions with Generative AI. The student will develop algorithms to detect sensitive information in user inputs and AI responses, implement methods to sanitize sensitive information before it reaches the AI model, and create user-friendly interfaces to alert users of potential privacy concerns.