PeopleofSuperintelligence

Researcher

Michael Kim

Also known as Michael P. Kim

Michael Kim is a computer scientist whose research develops algorithms for responsible machine learning. He earned a computer science PhD at Stanford in 2020 and served as a Miller postdoctoral fellow at UC Berkeley, hosted by Shafi Goldwasser, in March 2023. His papers cover multicalibration, auditing risk predictions, hidden model backdoors, privacy certification and omniprediction. As of October 2026, he is an assistant professor of computer science at Cornell, where he coauthors work on hybrid online learning.

Timeline

  1. 2018

    Multicalibration for fair predictions

    In 2018, he coauthors Multicalibration: Calibration for the (Computationally-Identifiable) Masses at ICML. The paper defines a fairness measure that offers calibrated predictions across computationally identifiable subgroups.

  2. 2020

    A Stanford PhD in fairness

    In 2020, he completes a computer science PhD at Stanford with the dissertation A Complexity-Theoretic Perspective on Fairness.

  3. November 2020

    Outcome Indistinguishability

    In November 2020, he coauthors Outcome Indistinguishability, a framework that uses computational indistinguishability to study prediction algorithms. The paper argues that auditors need oracle access to risk-prediction systems.

  4. 2022

    Undetectable backdoors in models

    In 2022, he coauthors Planting Undetectable Backdoors in Machine Learning Models. The paper presents one method using digital signatures and another for Random Fourier Features or Random ReLU models.

  5. March 2023

    A Miller postdoc at Berkeley

    In March 2023, he is a postdoctoral research fellow at UC Berkeley's Miller Institute, hosted by Shafi Goldwasser. His seminar presents research on foundations of responsible machine learning.

  6. 2024

    Certifying private mechanisms

    For CRYPTO 2024, he coauthors Certifying Private Probabilistic Mechanisms. Its protocol requires an untrusted party to follow a specified probabilistic mechanism or be caught, and the paper demonstrates certification for differentially private counting queries.

  7. 2025

    Algorithms for omniprediction

    In 2025, he coauthors Near-Optimal Algorithms for Omniprediction, which develops online and offline learning algorithms for predictions across a class of loss functions.

  8. As of October 2026

    Assistant professor at Cornell

    As of October 2026, he is an assistant professor of computer science at Cornell and researches algorithmic methods for responsible machine learning. In 2026, he coauthors Oracle-efficient Hybrid Learning with Constrained Adversaries, an efficient ERM-oracle algorithm for online learning with constrained adversarial labels.

Career

  1. NowCornell UniversityAssistant professor of computer science
  2. 2023University of California, BerkeleyPostdoctoral research fellow

Education

  1. 2020Stanford UniversityPhD

From the biography.

  1. 11% of AI doctorate holders from 2010 to 2016 founded or led a company or lab by 10 years after graduationIncludes Cornell University. We estimate how often AI doctorate holders founded or led a company or lab by 10 years after graduation, how waiting times compare between founders and top jobs, and which universities trained the most.
  2. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Cornell University. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  3. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorateIncludes Cornell University. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorate. Across 2,116 people with both an undergraduate degree and a doctorate recorded, we trace which undergraduate schools feed which doctoral programs, how often students stay at the same school, and how paths differ by birth-country region.
  4. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Cornell University. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.

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 Michael Kim yet. Send us a free-licensed photo.

  1. Multicalibration: Calibration for the (Computationally-Identifiable) Masses proceedings.mlr.press. Checked 21 October 2026.
  2. A complexity-theoretic perspective on fairness | Stanford Digital Repository purl.stanford.edu. Checked 21 October 2026.
  3. [2011.13426] Outcome Indistinguishability arxiv.org. Checked 21 October 2026.
  4. [2204.06974] Planting Undetectable Backdoors in Machine Learning Models arxiv.org. Checked 21 October 2026.
  5. CITP Seminar: Michael Kim – Foundations of Responsible Machine Learning | Princeton Alumni alumni.princeton.edu. Checked 21 October 2026.
  6. Certifying Private Probabilistic Mechanisms eprint.iacr.org. Checked 21 October 2026.
  7. [2501.17205] Near-Optimal Algorithms for Omniprediction arxiv.org. Checked 21 October 2026.
  8. Michael Kim | Department of Computer Science | Cornell Bowers cs.cornell.edu. Checked 21 October 2026.
  9. [2603.04546] Oracle-efficient Hybrid Learning with Constrained Adversaries arxiv.org. Checked 21 October 2026.