Her contribution

Joy Buolamwini

1989-

With Timnit Gebru, published Gender Shades in 2018, measuring errors in three commercial gender-classification systems across skin-type and gender groups.

Knowledge
Justice
Body
Recognition
Voice
Technology

Why this record matters

Impact and significance

The study found the highest error rates for darker-skinned women in the evaluated groups, showing why a single overall accuracy score can conceal disparities.

This is Matria’s interpretation, not a quotation or personalised advice.

Evidence notes

How we checked these statements

Each note connects one statement to the source and exact location Matria reviewed. It also names where the evidence stops.

What she did

With Timnit Gebru, Joy Buolamwini published Gender Shades in 2018, measuring errors in three commercial gender-classification systems across skin-type and gender groups.

Joy Buolamwini and Timnit Gebru. Gender Shades. PMLR 81:77-91, 2018.

Where to look: Abstract and pp. 77-91

What this source supports: direct

What this does not establish: The study evaluated three commercial systems and does not establish the performance of every facial-analysis system.

See this note’s review history

2026-09-06 · version 1 · source-inspected
The research paper and reported subgroup results were checked. Specialist historical review is not implied.

Why it mattered

Gender Shades found the highest error rates for darker-skinned women in the evaluated groups, showing why one overall accuracy score can conceal disparities.

Joy Buolamwini and Timnit Gebru. Gender Shades. PMLR 81:77-91, 2018.

Where to look: Abstract and pp. 77-91

What this source supports: direct

What this does not establish: The result is bounded to the systems, dataset and categories evaluated in the paper.

See this note’s review history

2026-09-06 · version 1 · source-inspected
The reported intersectional error comparison was checked against the paper. Specialist review is not implied.

Reviewing a source is not the same as specialist certification. Matria shows this trail so readers can inspect and challenge the wording.

Follow her work

Works in the archive

Gender Shades (2018, with Timnit Gebru)
Founded the Algorithmic Justice League (2016)
Unmasking AI (2023)

Read further

Sources and access

Joy Buolamwini and Timnit Gebru. Gender Shades. PMLR 81:77-91, 2018.
Abstract and pp. 77-91
Buolamwini, Joy. Unmasking AI: My Mission to Protect What Is Human in a World of Machines. Random House, 2023.Bibliographic reference. Consult the named edition through a library or publisher.
Buolamwini, Joy and Gebru, Timnit. Gender Shades. Proceedings of Machine Learning Research, 2018.Bibliographic reference. Consult the named edition through a library or publisher.

This label concerns the sources on this record. It does not validate every argument an author makes or establish what is true for a reader. Bibliographic coverage does not mean every passage or quotation has been verified. Specialist review remains pending unless a statement above says otherwise.

Find Matria on the App Store