The Washington PostDemocracy Dies in Darkness

‘60 Minutes’ ran an episode about algorithm bias. Only White experts were given airtime.

The episode renewed calls to #CiteBlackWomen, many of whom have been leading research on AI bias

By
June 11, 2021 at 3:22 p.m. EDT
(iStock; Washington Post Illustration/iStock; Washington Post illustration)

A recent “60 Minutes” episode on the use of facial recognition in policing has drummed up controversy in the form of a roughly 6,000-strong petition, an editor’s note from CBS — and a renewed call to #CiteBlackWomen.

The 13-minute-long segment, which aired May 16, reported on how facial recognition technologies have led to the wrongful arrests of Black men. It featured interviews with two White experts in facial recognition technologies as well as two Black men who were wrongfully arrested based on faulty facial recognition.

Joy Buolamwini, an artificial intelligence bias researcher at Massachusetts Institute of Technology and a Black woman, was not featured in the episode after spending what she said were between eight and 10 hours working with “60 Minutes” producers over the course of a few months, recommending research to incorporate and even building a custom demo program showing how facial recognition technologies analyze faces, she said.

It’s not just Sarah Milov. Female academics aren’t credited in media ‘all the time.’

Two days after the segment aired, CBS News released an editor’s note naming Buolamwini and her organization, Algorithmic Justice League — a digital advocacy organization that aims to raise awareness of the social implications of artificial intelligence — acknowledging her contributions to the research process, and noting that they faced time constraints that left them unable to mention the “dozens of sources — off and on camera — who helped us develop and focus this segment.” A CBS spokesperson directed The Lily to the editor’s note.

Now, Buolamwini is one of the Black women scholars leading the charge calling for change: A creator of the petition, Buolamwini characterizes the incident as part of a pervasive pattern of Black women not being credited for their research — both within artificial intelligence and beyond.

“In almost every facet of our lives, from technology to government and social movements, Black women are often assumed to be available and best equipped to do the hard work of moving an issue forward, and at the same time are not given the recognition for doing so,” said Buolamwini, whose TED Talk on algorithmic bias has 1.4 million views and whose research was featured in the Netflix documentary “Coded Bias.” “People act surprised when this happens, but exclusion, discrimination and bias in technology are the consequences of decisions like this being made every day. The result is AI and other technologies don’t work for Black women, people of color and other underrepresented communities.”

Buolamwini’s petition quickly gained traction on social media. It demands CBS News publicly apologize for the omission, institute a policy of formally crediting all sources who inform their productions and produce a segment by the end of this year focused on the Black women leading research to expose algorithmic harms — including Deborah Raji, who contributed to Buolamwini’s graduate research; Timnit Gebru, a former co-leader of Google’s Ethical Artificial Intelligence Team who said she was abruptly fired from Google last December for sending an email criticizing the company’s treatment of minority employees; and Tawana Petty, national organizing director for Data for Black Lives. The Lily could not reach Gebru and Petty by publication time.

A White city official refused to address this Black professor as ‘doctor.’ He got fired.

Black women have been at the forefront of leading research on the social and ethical implications of artificial intelligence — including University of California at Los Angeles professor Safiya Noble, author of “Algorithms of Oppression: How Search Engines Reinforce Racism” and Princeton University professor Ruha Benjamin, author of “Race After Technology: Abolitionist Tools for the New Jim Code.” But their research has repeatedly been overlooked and diminished by others in the tech industry, according to Buolamwini and Raji.

For Raji, her own identity as a Black woman has been crucial to her pursuit of research on gender- and race-based algorithmic biases — something she wishes others in the field took more seriously.

“It’s very frustrating to see a problem, vocalize a problem, and not be taken seriously,” she said. “The less we respect and pay attention to those impacted communities, the more likely these problems will continue to fester.”

But exclusion and lack of recognition of Black women scholars are not limited to the field of artificial intelligence, according to Christen Smith, an associate professor of anthropology and African and African diaspora studies at the University of Texas at Austin.

In 2017, Smith was sitting through a presentation at an academic conference when she noticed some of her own work paraphrased on a screen without any citation, she said.

She soon sprung into action, printing a T-shirt emblazoned with the sentence “Cite Black Women,” which she wore to a November 2017 national women’s studies conference. The incident was meant to “dare people not to cite me when they see me with that shirt on,” she said.

More than three years later, Cite Black Women has evolved into a movement that encourages professors, students and those outside university walls alike to acknowledge, amplify and read Black women’s work. Boulamwini included the hashtag #CiteBlackWomen in her petition, too.

Smith’s own research also looks at this phenomenon: A study she co-led whose results were published in the May issue of the journal Feminist Anthropology found that Black women anthropologists are severely underrepresented in top-tier journals, constituting only 0.87 percent of cited scholars — or 46 out of 5,445 — even though they make up 2.6 percent of American anthropologists overall. And most of those citations of Black women anthropologists’ work came from other Black authors, the study found.

For scholars, the amount of times their papers and books are cited are used to determine things like promotion, tenure and “deciding how much people are worth on the academic market,” Smith said.

But citation is also a practice grounded in respect and recognition, she added: “Citation is basically our way of acknowledging our genealogies of thought, and specifically it’s our way of acknowledging the intellectual labor that people put into their ideas, and the ways that those ideas reverberate beyond their immediate context.”

Recognizing Black women’s intellectual labor and giving them credit for their own ideas is also crucial outside of academia, considering the history of slavery and its contemporary reverberations, Smith said.

“Part of what’s really difficult and insidious is that there’s a whole tradition of Black people doing work and then White people taking the credit for it,” she said.

To Buolamwini, the Black women leading research on artificial intelligence deserve nothing less than full citations, a seat at the table and time on TV.

“AI continues to develop at a rapid pace, and centering the voices of Black women who are doing the work is key to understanding and addressing the ways AI fails marginalized identities,” she said.