Award-winning investigative journalist Julia Angwin is behind some of the most influential exposés on algorithmic bias. The editor in chief of The Markup, she has uncovered biases in algorithms used in pretrial risk assessments, insurance premium determinations, ride-sharing apps, social media advertising, and more. Celebrated for her ability to demystify complex algorithmic injustices and hold powerful corporations accountable, Angwin has shaped discourse around the disparate impact of algorithms by providing numerous striking and concrete examples.

Joy Buolamwini is the founder of the Algorithmic Justice League, an award-winning researcher, and poet of code. She advises world leaders, policymakers, and executives on redressing algorithmic harms. Her work is featured in global exhibitions and the documentary Coded Bias, available on Netflix. Her seminal research paper with Timnit Gebru, Gender Shades, exposed how poorly big-brand facial recognition systems perform on darker-skinned women's faces. The study's findings prompted swift action by Microsoft, IBM, and Amazon to address bias in their designs. Their continued advocacy has drawn massive public awareness to algorithmic biases in everyday technologies.

Virginia Eubanks' writing on technology and social justice highlights the effects of technological discrimination against working-class and poor communities. Her book, Automating Inequality: How High-Tech Tools Profile, Police, and Punish the Poor, is an eye-opening survey of how corporations and governments weaponize big data tools against the most marginalized people.

Professor Safiya Noble is the author of Algorithms of Oppression, which documents how Google search reinforces racial and gender stereotypes. She persuasively argues that the field needs a Black feminist theory for technology studies. Without widespread adoption of such a lens, daily-use technologies will continue reinforcing Black women and girls' marginalization.

Latanya Sweeney was the first scientist to demonstrate racial biases in commercial algorithms. She found that Google AdSense delivered ads that suggested an individual's arrest more often when searching for Black sounding names than for white-sounding ones. Sweeney found a Black-identifying name to be 25% more likely to get an ad suggestive of an arrest record. This early work established the field of algorithmic fairness.

Timnit Gebru is the founder and executive director of the Distributed Artificial Intelligence Research Institute (DAIR). Prior to that she was fired by Google in December 2020 for raising issues of discrimination in the workplace, where she was serving as co-lead of the Ethical AI research team. She received her PhD from Stanford University, and did a postdoc at Microsoft Research, New York City in the FATE (Fairness Accountability Transparency and Ethics in AI) group, where she studied algorithmic bias and the ethical implications underlying projects aiming to gain insights from data. Timnit also co-founded Black in AI, a nonprofit that works to increase the presence, inclusion, visibility and health of Black people in the field of AI, and is on the board of AddisCoder, a nonprofit dedicated to teaching algorithms and computer programming to Ethiopian highschool students, free of charge.

Cathy O’Neil

April Glasser

John Keefe

Jeremy Merrill

Karen Hao

The Markup Staff

Christian Sandvig

Kashmir Hill

Alexander Amini and his co-authors developed a technical solution to algorithmic bias to detect which training inputs come from marginalized groups (e.g., faces with dark skin). This method then samples from those inputs more frequently. This approach results in more balanced representations of outputs that approximate parity despite inequities in training sets.

The IBM team at AIF360 has put together one of the most comprehensive toolkits for detecting and mitigating algorithmic bias in various data science applications. Among their solutions, they offer data-scientist-oriented introductions to the concepts and techniques that undergird their toolkit, which is made free to the public.

Sarah Hajian works at the forefront of bias mitigation through data mining techniques. Her work has paved the way for reducing bias from unrepresentative training data sets by detecting and minimizing discrimination in the data collection and curation phases—otherwise known as data mining processes that are discrimination-conscious by design.

Weiwei Pan and Finale Doshi-Velez, two computer scientists at the intersection of machine learning and healthcare, have developed uncertainty-aware prediction models. Scientists can use this class of models to make predictions that are just as accurate as traditional models while also providing a metric for how confident a model is on any given prediction. In high-stakes tasks, such as medical diagnosis or judicial sentencing, uncertainty-aware methods can cue human decision-makers to the shortcomings of machine-learning models and mitigate real-world repercussions.

Brian Hu Zhang, Blake Lemoine, and Margaret Mitchell advanced adversarial debiasing, a popular debiasing tool used in development. In short, this method aims to learn a representation of the training set that is independent of some protected attribute (e.g., gender) by jointly training an adversary model to prevent the prediction of that same attribute.