Timnit Gebru, a high-profile AI ethics researcher and leader in the field, has been forced out of Google.
Gebru, who announced the news on Twitter Wednesday night, wrote that she was fired by Google’s head of AI Jeff Dean over an email she had written to a list consisting of women and allies at Google Brain. In addition, she shared that her manager’s manager sent an email to all of her direct reports saying that Gebru had resigned even though she had not done so. While Gebru did not respond to a request for comment by press time, her tweets indicate she is currently looking for an employment lawyer.
The email, which was obtained by Casey Newton of Platformer, lays out a situation that implies Gebru was going through the process of publishing a research paper when she was told that she needed to retract the paper with little explanation, conversation, or opportunity to address concerns. The email expresses Gebru’s frustrations with the lack of transparency in the review process and the way she was treated when she asked for more information. In her tweets about the situation, she says that Google told her the email was “inconsistent with the expectations of a Google manager.” You can read the full text of Gebru’s email here.
According to a source, the paper in question was critical of big language models, of which Google is heavily invested in building. In particular, the paper critiques these models from an ethical and environmental perspective.
Google did not respond to a request for comment, but pointed to an email also published in Platformer that Dean sent to employees Thursday morning. In the email, he wrote that Gebru’s paper did not meet the company’s criteria for publication because it lacked reference to recent research. In addition, he wrote that Gebru said “she would leave Google and work on an end date” unless a set of conditions were met, including that she learned the “exact feedback” to the paper along with the names of everyone who provided comments. Dean’s email did not address why Gebru was immediately locked out of her email, why her team was told she had quit without a formal resignation or established end date, or how the email she sent factored into this treatment.
Gebru is a widely respected researcher who also cofounded the nonprofit Black in AI to increase representation in a heavily white field. She is perhaps best known for her contributions to a landmark paper in 2018 that showed how facial recognition systems had hugely different error rates for images of white men as compared to Black women.
Gebru was fired on the same day that the National Labor Relations Board accused Google of violating U.S. labor law for allegedly firing two employees over protesting Google policies and attempting to organize. She joins a growing group of former Google employees who say they have been pushed out in retaliation for speaking out over the last few years.
At the Fast Company Innovation Festival in October, Gebru spoke to my colleague Amy Farley about the challenges of pushing back against entrenched interests. “You can’t set up a system where the only incentive is to make more money and then just assume that people are going to magically be ethical,” she said.
Earlier this week, Gebru inquired on Twitter about whether there is any regulation in the works to protect ethical AI researchers, similar to how there is legislation that safeguards whistleblowers.
The outpouring of support for Gebru was immediate and overwhelming. Members of her team and her direct reports tweeted in solidarity, as did other AI researchers, activists, and tech workers from across the industry.
To be bold, to be honest, to know your worth, to speak up for others, to do the thankless work, to pioneer the way, to suffer no fools, to be all they could only wish to be, wrapped in black womanhood is an affront to those who would rather see us cower than own our power 1/n
— Joy Buolamwini (@jovialjoy) December 3, 2020
— Mark Díaz (@blahtino) December 3, 2020
I thought this was a joke because it seemed ridiculous that anyone would fire @timnitGebru given her expertise, her skills, her influence. This is one of the many times when I think there is just no hope for the tech industry. https://t.co/2Px7nkObke
— Ellen K. Pao (@ekp) December 3, 2020
The amount of disrespect and straight-up gaslighting that @timnitGebru is at the receiving end from our employer right now is chilling.
Google could be one of the research leaders on the social implications of AI and tech. But they're choosing to go the other way.
— Dr. Alex Hanna (@alexhanna) December 3, 2020
1) Timnit is easily one of the sharpest, clearest-minded people I’ve ever interacted with, however briefly
2) judge companies and management by what they do, not what they say
3) go look at the rest of timnit’s team reactions
This fuckup by google is monumental https://t.co/FxcyVOekiP
— Carlos Scheidegger (@scheidegger) December 3, 2020
Just like that, Google lost the little legitimacy it had. @timnitGebru has been nothing but the ideal role model for Black women, an aspiration, and a towering figure that keeps pushing the whole field of AI ethics to respectable standard. https://t.co/LfMc8XuSOH
— Abeba Birhane (@Abebab) December 3, 2020
This story has been updated.