PeopleofSuperintelligence

Researcher

Aleksander Mądry

Aleksander Mądry is a computer scientist whose work connects graph algorithms with robust machine learning. He studied physics and computer science at the University of Wrocław, completed a computer science PhD at MIT in 2011, and joined MIT EECS as an assistant professor in February 2015. At EPFL, he headed Theory of Computation Laboratory 1 and won FOCS Best Paper in 2013 for maximum-flow work; his 2017 paper applies robust optimization to adversarial attacks. He received the Presburger Award in 2018, coauthored a 2019 paper on non-robust data features, helped establish MIT's Center for Deployable Machine Learning, and co-led MIT's AI Policy Forum. OpenAI named him to lead Preparedness in October 2023 and moved him to a reasoning-model team in July 2024; after leaving, he announced plans in October 2026 to return and work with Jakub Pachocki on the pace and implications of AI progress and their effects on strategy, and as of October 2026 he is a Visiting Scientist in MIT EECS and a Temporary Visiting Professor at MIT CSAIL.

Now at
MIT CSAIL
Works
141
Citations
14,444
h-index
43
Badges

Publication numbers from OpenAlex, as of 27 October 2026.

Timeline

  1. By 2011

    Wrocław and an MIT PhD

    Born in Wrocław, Poland, he studies physics and computer science at the University of Wrocław in the mid-2000s. At MIT, he works with advisors Michel X. Goemans and Jonathan A. Kelner and earns a computer science PhD in 2011.

  2. 2013

    Maximum flow at EPFL

    After a year at Microsoft Research New England, he heads EPFL's Theory of Computation Laboratory 1. In 2013, he wins FOCS Best Paper for an algorithm that improves maximum-flow bounds.

  3. February 2015

    Assistant professor at MIT

    In February 2015, he joins MIT EECS as an assistant professor. His research tackles fundamental algorithmic problems motivated by real-world optimization.

  4. June 2017

    Training robust networks

    In June 2017, he coauthors Towards Deep Learning Models Resistant to Adversarial Attacks with Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras and Adrian Vladu. The paper studies adversarial robustness through robust optimization and develops methods for training and attacking neural networks.

  5. 2018–2019

    A Presburger Award and feature paper

    He receives the Presburger Award in 2018 for his contributions to algorithm theory, including maximum-flow work. In May 2019, he coauthors Adversarial Examples Are Not Bugs, They Are Features, which links adversarial examples to highly predictive, non-robust data features.

  6. By 2020

    Deployable machine learning at MIT

    By 2020, he and his colleagues establish the MIT Center for Deployable Machine Learning within the MIT Quest for Intelligence. The center develops machine-learning tools for real-world deployment.

  7. 2022

    AI Policy Forum co-lead

    In 2022, he co-leads MIT's AI Policy Forum and closes its symposium with a call for researchers, industry, policymakers and other stakeholders to work together.

  8. October 2023

    OpenAI's Preparedness team

    In October 2023, OpenAI announces its Preparedness team with him as its lead to assess catastrophic risks from AI models.

  9. July 2024

    Reasoning work at OpenAI

    In July 2024, OpenAI reassigns him to a team working on reasoning models.

  10. As of October 2026

    OpenAI return and MIT roles

    After leaving OpenAI, he announces plans in October 2026 to return and work with OpenAI scientist Jakub Pachocki on the pace and implications of AI progress and their effects on strategy. He is a Visiting Scientist in MIT EECS and a Temporary Visiting Professor at MIT CSAIL. At CSAIL, he leads the Reliable and Robust Machine Learning project.

Career

  1. NowMassachusetts Institute of TechnologyVisiting Scientist
  2. NowMIT CSAILTemporary Visiting Professor
  3. NowMIT CSAILLead, Reliable and Robust Machine Learning project
  4. From 2024OpenAIReasoning-model team member
  5. From 2023OpenAIPreparedness Lead
  6. From 2022Massachusetts Institute of TechnologyCo-lead of AI Policy Forum
  7. From 2015Massachusetts Institute of TechnologyAssistant Professor
  8. EPFLHead of Theory of Computation Laboratory 1
  9. Massachusetts Institute of TechnologyCo-founder of MIT Center for Deployable Machine Learning

Education

  1. 2011Massachusetts Institute of TechnologyPhD
  2. University of Wrocław

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Checked career sources: csail.mit.edu.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

Photo: We have no published photo of Aleksander Mądry yet. Send us a free-licensed photo.

  1. “Doing machine learning the right way” | MIT News | Massachusetts Institute of Technology news.mit.edu. Checked 27 October 2026.
  2. New faculty in engineering | MIT News | Massachusetts Institute of Technology news.mit.edu. Checked 27 October 2026.
  3. Aleksander Madry receives Best Paper award at FOCS 2013 - EPFL actu.epfl.ch. Checked 27 October 2026.
  4. [1706.06083] Towards Deep Learning Models Resistant to Adversarial Attacks arxiv.org. Checked 27 October 2026.
  5. Aleksander Madry wins the Presburger Award | MIT CSAIL Theory of Computation toc.csail.mit.edu. Checked 27 October 2026.
  6. [1905.02175] Adversarial Examples Are Not Bugs, They Are Features arxiv.org. Checked 27 October 2026.
  7. Exploring emerging topics in artificial intelligence policy | MIT News | Massachusetts Institute of Technology news.mit.edu. Checked 27 October 2026.
  8. OpenAI forms team to study 'catastrophic' AI risks, including nuclear threats | TechCrunch techcrunch.com. Checked 27 October 2026.
  9. Aleksander Mądry — Wikipédia fr.wikipedia.org. Checked 27 October 2026.
  10. “Prioridad central”: un exjefe de Seguridad de OpenAI vuelve a la compañía ante la creciente preocupación por el avance de la IA - LA NACION lanacion.com.ar. Checked 27 October 2026.
  11. Aleksander Mądry - MIT EECS eecs.mit.edu. Checked 27 October 2026.
  12. Aleksander Madry | MIT CSAIL csail.mit.edu. Checked 27 October 2026.