Data Security and Privacy in the Age of AI (CS 5914) 
 
    Time and Location :  M-W: 4pm-5:30pm   @D&DS 240

     
       Instructor                            :   Murat Kantarcioglu
     Office Hours & Location   :   Monday: 5:30pm to 6:30pm or by appointment on online
 
 
     Teaching Assistant             :   TBD
       

Prerequisites                      :   Basic knowledge of databases and machine learning
                       

Grading:

  •   Homework   % 10 (4 homework, each worth 2.5%)
  •   Project          % 40 (Group project that requires programming)
  •   Midterm       % 25
  •   Final             % 25

 

 
Course Topics: 

This course provides essential tools for safeguarding sensitive data in AI applications while ensuring robust privacy protections. 
Topics covered include: Fundamentals of Access Control, encryption-based methods for secure data processing, 
confidential computing, blockchain for enhancing data integrity, data anonymization, differential privacy, 
and approaches to understanding and defending against privacy and security threats to AI models.. 

Learning outcomes:
  • Ability to understand and use basic cryptographic techniques and tools for data security and privacy
  • Ability to understand and use access controls for AI applications
  • Ability to understand and use integrity policies
  • Ability to understand and use database access control tools
  • Ability to understand and use defensive tools against common data management system cyber attacks
  • Ability to understand and use basic privacy-enhancing  AI technologies.
  • Ability to understand and use defensive tools for protecting AI systems from various attacks.

         
           Textbook:   None.

 Course Outline:

Week 1: 01.20.25
  • Overview of the Data Security and Privacy for AI
  • Design principles for Security
  • Access control basics
  • Reading:  Fred B. Schneider’s book chapter  (pdf)

Week 2:
01.27.25
  • Access Control Foundations
  • Reading: Fred B. Schneider’s book chapter (pdf)
  • Reading:  HRU paper (pdf)
  • Reading:  NIST Attribute Access Control Model (Till Section 3) (pdf)

Week 3:
02.03.25
  • Access control models
  • Integrity/Hybrid Models

Week 4:
02.10.25
  • Basic Cryptography Overview
  • Authentication
  • Reading:  Fred B. Schneider’s book chapter (pdf)
  • Homework 1 is available.
  • Project Description is available.
Week 5:
02.17.25

Week 6:
02.24.25
  • Database Security
  • Encrypted Data storage in Databases
  • Reading:  Please read the following overview paper (pdf)
  • Reading:  Please read the following tutorial from Microsoft Research (pdf)

Week 7:
03.03.25
  • Confidential Computing
  • Reading:  Intel Sgx Overview (link) 
  • Reading:  Nvidia Confidential Computing (link)
Week 8:
03.10.25
  • SPRING BREAK

Week 9:
03.17.25
  • Midterm review (03.17.25)
  • !!! MIDTERM (03.19.25, Thursday) in class. !!!

Week 10:
03.24.25
  • SQL and Code injection attacks
  • Reading: Please see the tutorial from Oracle.
  • Homework 2 is available on elearning

Week 11:
03.31.25
  • Introduction to Data Privacy
  • Reading: K-annonymity (pdf), differential-privacy (pdf),

Week 12:
04.07.23
  • Introduction to Data Privacy cont.
  • Homework 3 is available on elearning

Week 13:
04.14.25
  • Introduction to Data Privacy cont.
  • Introduction to Federated Learning
  • Reading: Privacy-preserving distributed data mining (pdf)

Week 14:
04.21.25
  • Attacks against Machine Learning / AI Models
  • Evasion Attacks
  • Poisoning Attacks
  • Reading: TBD

Week 15:
04.28.25
  • Attacks against Large Language Models
  • Compliance for AI/ML
  • Reading: TBD
Week 16
05.05.25
  • Combining multiple techniques for ML/AI Security and Privacy
  • Project Presentations.

????
  • Final EXAM