Safety of GenAI Through the Lens of Security and Cryptography

סקירה כללית

סילבוס

• Introduction to Generative AI
Overview of Generative AI
Recent advancements and applications
Importance of safety in Generative AI
• Security Challenges in Generative AI
Identifying potential threats
Privacy concerns and data protection
Adversarial attacks on Generative AI models
• Cryptographic Techniques for AI Safety
Basics of cryptography relevant to AI
Encryption methods to secure AI models
Privacy-preserving machine learning techniques
• Somesh Jha's Theoretical Approach
Overview of Jha's contributions to AI safety
Insights from theoretical frameworks
Case studies and practical examples
• Trustworthy AI Systems
Principles of designing trustworthy AI
Verification and validation of AI models
Building resilient and robust Generative AI
• Ethical Considerations in AI Security
Balancing privacy and functionality
AI governance and regulatory frameworks
Societal impacts and ethical dilemmas
• Practical Approaches and Tools
Tools for securing Generative AI
Implementing cryptographic solutions
Hands-on exercises and projects
• Future Directions for AI Safety
Emerging trends in AI security and cryptography
Research opportunities and open challenges
Collaboration between academia and industry
• Conclusion and Review
Summary of key takeaways
Final discussions and Q&A session
Course feedback and evaluation

---

נושאים

Computer Science