Piotr (Peter) Mardziel

CMU Silicon Valley
Office: B23 114

Piotr (Peter) Mardziel

About me

I'm a systems scientist at CMU where I work on transparency and accountability in machine learning with applications to privacy and fairness.

Previously I was a PhD student and post-doc in the Programming Languages at University of Maryland (PLUM) group where I worked on probabilistic programming with applications for privacy and quantitative information flow. My interests include:

S&P Topics:
  • Anonymity
  • Language-based security
  • Machine learning and AI security
  • Privacy technologies and mechanisms
  • Secure information flow
CSF Topics:
  • accountability
  • anonymity and privacy
  • formal methods and verification
  • information flow control
  • language-based security
  • machine learning
  • software security
CMU Keywords:
  • Computer Systems: Security and Privacy:
    Foundations and Formal Methods, Analytics and Machine Learning
  • Signal Processing, Communications and Controls: Theoretical Foundations:
    Information Theory
  • Computer Systems: Application Domains:
    Machine learning
  • Signal Processing, Communications and Controls: Data Science:
    Machine Learning, Deep Learning


  • Foundations of Privacy
    CMU ECE 18-734 Fall 2018
  • Privacy and fairness for machine learning
    with Anupam Datta, Matt Fredriskon, students Shayak Sen, Gihyuk Ko
  • Transparency and interpretable machine learning
    with Anupam Datta, students Sophia Kovalova, Raj Gupta
  • Static analysis, abstract interpretation of java bytecode
    with PLUM


  • Privacy with origin
    with Anupam Datta, Michael Tschantz, Sebastian Benthall, Helen Nissenbaum
  • Quantitative information flow for dynamic data
    with Mario Alvim, Michael Hicks, Michael Clarkson
  • Game theoretic analysis of password authentication
    with MHR Khouzani, Carlos Cid, Mudhakar Srivatsa
  • Inference control
    with Michael Hicks, Stephen Magill, Mudhakar Srivatsa, Jon Katz
  • Abstract interpretation of probabilistic program semantics
    with Michael Hicks, Stephen Magill
  • Secure software development contest
    Builid It, Break It, Fix It
    with PLUM


Other service


Tech-reports / Pre-prints