Top AI ethics researcher says Google fired her; the company denies it

OAKLAND, Calif. (Reuters) – A top Google scientist on ethical artificial intelligence says he was fired after criticizing the company’s diversity efforts, a claim disputed by Alphabet Inc on Thursday in the last interview between Internet giant and worker activists.

Timnit Gebru, who is black, said on Twitter that she was fired on Wednesday after sending an email to colleagues expressing frustration with the gender diversity within Google’s AI unit and wondering if company executives scrutinized his work more rigorously than that of people from different backgrounds. Gebru co-founded the non-profit organization Black in AI which aims to increase the representation of people of color in artificial intelligence and co-authored a landmark article on bias in facial analysis technology.

Jeff Dean, head of Google’s AI unit, told staff in an email reviewed by Reuters that Gebru had threatened to resign unless he was told which colleagues believed a draft document she had written was unpublishable, a request rejected by Dean.

“We accept and respect his decision to quit Google,” Dean wrote in the email, adding, “We all genuinely share Timnit’s passion for making AI more equitable and inclusive.”

Gebru said in a series of Twitter posts that Google cut her off from its systems without warning or talking to her about her concerns.

Gebru’s abrupt departure comes on top of years of angst, including several resignations and layoffs, in the AI ​​department and other Google organizations because of diversity and whether the efforts of the company to minimize the potential damage of its services are sufficient.

More than 150 employees have expressed support for Gebru, demanding that Google step up its commitment to academic freedom and explain why it has chosen to “censor” its newspaper, according to a petition posted online. Sherrilyn Ifill, president of the Legal Defense and Educational Fund of the NAACP, wrote on Twitter that Gebru’s dismissal was “absolutely exasperating” and “a disaster.”

Just on Wednesday, the National Labor Relations Board filed a complaint accusing Google of illegally monitoring and questioning several workers who were later fired for protesting company policies and trying to organize a union.

Gebru’s article argued that tech companies could do more to ensure that AI systems aimed at mimicking human writing and speech don’t exacerbate historical gender biases and the use of offensive language. , according to a draft copy seen by Reuters.

In his email to staff, Dean said the article was not handed over to the company for timely review and was submitted to a conference without Google’s permission.

He also challenged some of his findings, which he said were based on outdated concerns, including the environmental impact of large numbers of computers processing data.

Responding to the company’s rejection of his work, Gebru wrote on Twitter last week: “Nothing like a group of privileged white men trying to crush research conducted by marginalized communities for marginalized communities in their own right. ordering to STOP with a ZERO conversation. The lack of respect is incredible. “

Google declined to comment on his departure beyond Dean’s email, which was first reported by tech news site Platformer.

Gebru previously worked at Microsoft Research and co-authored a widely cited 2018 article that revealed higher error rates in facial scanning technology for women with darker skin.

His new article, co-authored with non-Google staff, is still expected to be featured at an IT conference in March, according to a person familiar with the matter.