Unmasking AI

Unmasking AI: My Mission to Protect What Is Human in a World of Machines is a memoir-and-manifesto book by Joy Buolamwini, published by Random House on October 31, 2023. The book traces Buolamwini's path from her early experience of being unrecognised by face-detection software at the MIT Media Lab — the moment she has described as the founding of the Algorithmic Justice League — through the Gender Shades research she undertook with Timnit Gebru, the policy advocacy that followed, and the formation of AJL as a research-and-organising vehicle for what she calls "algorithmic justice."

The book makes a sustained case for treating AI as a sociotechnical system in which the absence of marginalised perspectives in design, training data, and audit produces concrete harms — and it argues that the people most likely to be misclassified, mis-served, or surveilled by AI systems are best positioned to identify those harms. Unmasking AI gives prominent treatment to two method concepts the book has helped to popularise: the coded gaze, the patterns of encoded discrimination and exclusion that result when narrow groups of designers and datasets shape the systems most of the public interacts with, and the evocative audit, an auditing practice that pairs the quantitative measurement of algorithmic harm with first-person testimony from the people the systems misclassify, mis-serve, or surveill. Specific cases the book builds on include the Gender Shades commercial-vendor audits that Buolamwini and Gebru published in 2018 and the Coded Bias documentary in which Buolamwini was the main subject. The closing chapters set out an advocacy programme — limits on facial recognition in policing, audits and red-teaming for high-risk systems, community-reporting infrastructure for those harmed by automated systems — that aligns directly with AJL's own programme priorities.

Within the corpus, Unmasking AI is the canonical book-length artefact for the Algorithmic Justice League: readers who follow AJL's public framings, programme names, or method concepts will frequently meet Unmasking AI as the source. The book has been a national bestseller and is widely cited in journalism and academic writing on AI accountability; alongside the Gender Shades paper and the Coded Bias documentary, it has been one of the principal vehicles through which the algorithmic-bias and algorithmic-justice frame has reached general audiences in the years since 2018.