
Machine learning and ethics researcher at Hugging Face
A researcher in algorithmic bias and fairness who introduced model cards for transparent model reporting and co-founded Google's Ethical AI team.
Margaret Mitchell is a computer scientist who works on algorithmic bias and fairness in machine learning. She earned a bachelor's degree in linguistics from Reed College in 2005, a master's in computational linguistics from the University of Washington in 2009, and a PhD from the University of Aberdeen in 2013 with a thesis on generating reference to visible objects.
She worked at Johns Hopkins University and then Microsoft Research from 2013, where she led the Seeing AI project, an app that narrates text and images for people who are visually impaired. In 2016 she joined Google Research, where she founded and co-led the Ethical AI team with Timnit Gebru. She is best known for introducing model cards for transparent reporting of what a model is intended for, and for debiasing methods based on adversarial learning. Google fired her in 2021 after an investigation following Gebru's dismissal, and she joined Hugging Face in late 2021.
Co-authored the model cards framework for transparent reporting of a machine learning model's intended use and performance across groups.
Founded and co-led the Ethical Artificial Intelligence team at Google Research with Timnit Gebru.
Was research lead of Seeing AI at Microsoft, an app that supports visually impaired people by narrating text and images.
Senior research scientist, co-founder of the Ethical AI team
Research lead, Seeing AI