"The computer scientist who wants ethics built into the math, not bolted on after"
Michael Kearns is a computer scientist and the National Center Chair Professor in the Computer and Information Science Department at the University of Pennsylvania, with a secondary appointment in Statistics and Data Science at the Wharton School. He is a pioneer of computational learning theory — the mathematical study of what machines can and cannot learn — and has spent decades at the intersection of machine learning, algorithmic game theory, and quantitative finance, including senior technical roles at Bell Labs, SoftBank, and Morgan Stanley. Kearns founded Penn's Networked and Social Systems Engineering program and its Warren Center for Network and Data Sciences. With longtime collaborator Aaron Roth, he co-authored The Ethical Algorithm, making the case that fairness and privacy can be engineered into code with the same rigor used to guarantee an algorithm's speed or accuracy.
Our in-depth summaries and reviews of his work
Kearns and Roth show how to build fairness and privacy directly into algorithms with differential privacy and formal fairness constraints, replacing after-the-fact…
Read Summary →Fairness, privacy, and accountability can be engineered into an algorithm's math — not bolted on as an afterthought.
When many algorithms interact in markets, pricing, or bidding, their collective behavior can look coordinated even when it isn't.
Some ethical tradeoffs have provable technical answers; others still require human judgment and public debate.
"Algorithms don't have to just be fixed after the fact by laws and lawsuits — they can be designed to behave well from the start."— The Ethical Algorithm
"When independent algorithms interact at scale, they can produce outcomes that look like collusion — without anyone coordinating anything."— The Ethical Algorithm
Kearns and Roth show how to build fairness and privacy directly into algorithms with differential privacy and formal fairness constraints, replacing after-the-fact policy patches with provable, mathematical guarantees.
