Algorithmic Justice League — publication: Gender Shades study (2018) by Buolamwini and Gebru demonstrated intersectional accuracy disparities in commercial facial recognition from IBM, Microsoft, and Face++. Dark-skinned women had highest error rates; light-skinned men had lowest. Over 4,900 citations on Semantic Scholar.

NoteThe provided source text is a JavaScript error page that does not contain any substantive information about the Gender Shades study, its authors, findings, citation count, or impact. While the URL appears to be the correct Semantic Scholar page for the study, the actual content is inaccessible in the provided excerpt. Without access to the actual source material, none of the specific claims (study details, citation count as of March 2026, IBM response, industry impact) can be verified or contradicted. This is a technical access issue rather than a content contradiction.

NoteThe provided source text is a JavaScript verification page, not the actual Semantic Scholar paper content. It contains no information about the Gender Shades study, the researchers (Buolamwini and Gebru), the findings regarding intersectional accuracy disparities, the specific error rates for different demographic groups, or the citation count. While the URL appears to point to the correct paper, the actual source content provided does not address any aspect of the claim. Without access to the actual paper or metadata, the claim cannot be verified against the source.