AI Safety, Security, and Privacy

This unit examines the risks, protections, and governance practices that make AI systems trustworthy, focusing on safety, security, privacy, and accountability. Students build on foundational AI concepts to analyze incidents, apply technical and organizational controls, and evaluate legal and ethical frameworks to prepare for advanced implementation and oversight.

Learning Objectives

• Analyze real-world and hypothetical incidents of compromised AI to identify technical failures, organizational weaknesses, and potential impacts on public safety and civil liberties.
• Apply leading security and technical-safety practices (e.g., threat modeling, robust testing, access controls, and incident response) to design a practical mitigation plan for an AI system.
• Demonstrate methods to assess and mitigate data-privacy risks—addressing limits of anonymization, consent models, access controls, and mapping controls to applicable laws and regulations.
• Evaluate accountability and governance frameworks by creating a stakeholder roles-and-responsibilities plan that includes auditing, reporting, and corrective-action procedures for AI-related harms or noncompliance.