Google is reorganizing its responsible AI teams in the wake of Timnit Gebru’s firing. The ethical AI team will now roll up to Marian Croak, a prominent Black executive in the engineering department. Croak will also oversee employees focused on engineering fairness products, according to Bloomberg. She will report to Jeff Dean, who leads the company’s AI efforts.
The ethical AI team was not aware of the reorganization until news broke Wednesday night.
In a blog post confirming Croak’s appointment, Google said the executive will be leading “a new center of expertise on responsible AI within Google Research.”
The change is an attempt to stabilize the department, which has been in turmoil for months, Bloomberg reports. In December, Timnit Gebru, co-lead of the ethical AI team, announced she’d been abruptly fired. The following month, the company began investigating her counterpart Margaret Mitchell, who had been using a script to go through her emails to look for examples of discrimination against Gebru. Mitchell now says she’s been locked out of her corporate accounts for more than five weeks.
Prior to her dismissal, Gebru had been trying to publish a paper on the dangers of large language processing models. Megan Kacholia, vice president of Google Research, asked her to retract the paper. Gebru pushed back, saying the company needed to be more transparent about the publication process. Shortly afterward, she was fired.
The ethical AI team published a six-page letter in the wake of Gebru’s termination, calling on Kacholia to be replaced. “We have lost confidence in Megan Kacholia and we call for her to be removed from our reporting chain,” the letter read.
Now, the team may be getting its wish. As part of the reorganization, Kacholia will no longer lead the ethical AI researchers, according to Bloomberg. It’s not clear what this means for Margaret Mitchell, who is still being investigated by the company.