Google is changing its search and diversity-related policies after completing an internal investigation into the dismissal of co-leader of the ethical AI team Timnit Gebru, according to Axios. The company intends to link the remuneration of certain executives to diversity and inclusion goals. It is also changing the way confidential employee exits are managed.
Although Google did not disclose the results of the investigation, the changes appear to be direct responses to how the situation with Gebru fell. After Google demanded that a co-authored article be withdrawn, Gebru told management of the research team that he would resign and work on a transition plan, unless certain conditions were met. Instead of a transition plan, the company immediately ended its job while she was on vacation. This sparked backlash from members of his team and even caused some Google engineers to withdraw in protest.
Google claimed that Gebru’s article was not submitted correctly, although the research team disagreed. Google has now said it will “streamline its research publication process,” according to Axios, but the exact details of the policy changes were not provided.
In an internal email to the team, Jeff Dean, head of AI at Google, wrote:
I heard and recognized what Dr. Gebru’s departure meant for women technologists, for those in the black community and other underrepresented groups looking for careers in technology, and for many who care deeply about the responsible use of AI for Google. This has led some to question their place here, which I regret.
He also apologized for the way Gebru’s departure was handled, although he has stopped calling it a dismissal.
Policy changes occur a day after Google restructured its AI teams, a change of which the members of the AI ethics team were “the last to know”, according to research scientist Alex Hanna, who is part of the team.
Google has refused to share updated policies with The Verge, instead of pointing to Axiosfor details.