Two of Google's Ethical AI Staffers Leave to Join Ousted Colleague’s Institute

(Bloomberg) -- Google's Ethical AI research group lost two more employees, adding to the turmoil at the unit studying an area that is vitally important to the technology giant's business future and political standing.

Alex Hanna, a research scientist, and Dylan Baker, a software engineer, resigned from the Alphabet Inc. unit on Wednesday to join Timnit Gebru's new nonprofit research institute, they said in an interview. The organization -- called DAIR, or Distributed AI Research -- launched in December with the goal of amplifying diverse points of view and preventing harm in artificial intelligence.

Hanna and Baker said they now believe they can do more good outside Google than within it.

There's work to be done “on the outside in civil society and movement organizations who are pushing platforms,” said Hanna, who will be DAIR's director of research. “And staying on the inside is super tiring.”

A Google spokesperson said in a statement, “We appreciate Alex and Dylan's contributions -- our research on responsible AI is incredibly important, and we're continuing to expand our work in this area in keeping with our AI Principles. We're also committed to building a company where people of different views, backgrounds and experiences can do their best work and show up for one another.”

Google's Ethical AI group has been roiled by controversy since 2020, when Gebru -- co-head of the team -- began speaking out about the company's treatment of women and Black employees. In December of that year, management dismissed Gebru (she said she was fired, while the company said it accepted her resignation) after a dispute over a paper critical of large AI models, including ones developed by Google.

Alphabet Chief Executive Officer Sundar Pichai apologized for how the matter was handled and launched an investigation, but it didn't quell the upheaval. Two months later, the company fired Gebru's co-head of Ethical AI research and one of the paper's co-authors, Margaret Mitchell, raising questions about whether researchers were free to conduct independent work.

A major concern was that data with biases is used to train AI models. Gebru and her co-authors expressed concern that these models could contribute to “substantial harms,” including wrongful arrests and the increased spread of extremist ideology.

The dismissals have weighed heavily on Hanna and Baker for the last year and staying at Google became untenable, they said. The two employees also said they wanted the opportunity to work with Gebru again.

In a resignation letter, Hanna said she believed Google's products were continuing to do harm to marginalized groups and that executives responded to those concerns with either nonchalance or hostility.

“Google's toxic problems are no mystery to anyone who's been there for more than a few months, or who have been following the tech news with a critical eye,” Hanna wrote. “Many folks -- especially Black women like April Curley and Timnit -- have made clear just how deep the rot is in the institution.” Google's researchers do good work “in spite of Google,” not because of it, she added.

Hanna and Baker have been vocal on issues such as workers' rights and military contracts at the tech giant, and they said the company seems more impervious to employee activism and public embarrassment than it was a few years ago.

They believe Google's high-profile 2019 firing of several activist employees had a chilling effect on workplace activism, paving the way for more controversial corporate decisions, such as an ongoing plan to pitch Google Cloud's services to the U.S. military.

A National Labor Relations Board judge is currently considering a complaint about the firings issued by agency prosecutors against the company, which has denied wrongdoing.

Baker, who will become a researcher at DAIR, said he's excited to be “able to do more work in the direction of building the kind of world that we want, which is equally as important as identifying harms that exist.”