AI-Driven Deception and Synthetic Threats

This unit examines how advanced AI models and synthetic media enable deceptive communication and cyber-physical attacks, including LLMs, transformers, deepfakes, voice synthesis, and robotic impersonation. Students will practice prompt engineering, conversational relay techniques, simulation-driven optimization of social engineering, and detection/mitigation strategies while assessing ethical, legal, and technical implications.

Learning Objectives

• Analyze transformer and large language model architectures, failure modes, and attack vectors to identify how they can be exploited for conversational deception and synthetic media generation
• Apply prompt engineering, context-crafting, conversational relays, and Monte Carlo simulation methods to design and quantify social engineering campaigns and detection risks
• Demonstrate hands-on techniques to create, detect, and mitigate synthetic content (deepfake imagery, voice duplication, and AI-generated text) and assess technical exploitation versus unintended problems
• Evaluate ethical, legal, and defensive countermeasures for AI-enabled deception—including staying below detection thresholds, incident response, and sociopolitical impact—when developing mitigation strategies