Google AI Ethics Leader Fired After Email Controversy

Timnit Gebru, a prominent researcher and influential figure in the areas of ethics and artificial intelligence, states that Google terminated her employment following an email she distributed to her team members. Gebru asserts that Google’s decision to end her employment stemmed from an email sent to her direct reports, which the company characterized as demonstrating “conduct not aligned with the standards expected of a Google manager.”

Gebru, who served as a co-leader of Google’s Ethical Artificial Intelligence team, shared her account on Twitter late last night, coinciding with the National Labor Relations Board’s filing of a complaint against Google alleging employee monitoring and unlawful dismissals.

Gebru explains that she was not directly informed of her dismissal, but instead received an email from a member of her management’s team stating:

“Thank you for clarifying your requirements. We are unable to accommodate requests #1 and #2. We acknowledge your decision to depart from Google as a consequence, and we are accepting your resignation.”

According to Gebru, the email further indicated that “specific elements of the email you sent last night to non-managerial staff within the brain group were deemed inconsistent with the expectations of a Google manager.”

The email in question, which was reviewed by Casey Newton, detailed Gebru’s dissatisfaction with her organization’s hiring practices, noting that despite discussions about diversity, only approximately 14% of new hires this year were women. She highlighted that Samy Bengio, a leader within the Google Brain team, achieved a 39% female hiring rate, but observed a lack of incentives for replicating this success. She also added:

Gebru’s communication also addressed concerns regarding the suppression of underrepresented perspectives, instances where her expertise was disregarded, and feelings of being manipulated while at Google.

We have contacted both Gebru and Google for their perspectives on this matter.

As reported by Bloomberg, Gebru has consistently voiced concerns about the limited diversity within the technology sector and the challenges faced by Black professionals in the field. Bloomberg reports that Gebru believes Google’s actions were intended to discourage other employees from speaking out.

Gebru is a recognized authority in the fields of ethics and artificial intelligence. In 2018, she collaborated with Joy Buolamwini, the founder of the Algorithmic Justice League, on a study examining biases in facial recognition technology. Their research revealed significant discrepancies in error rates between individuals with lighter skin tones and those with darker skin tones, leading to the conclusion that these systems performed inadequately for people with darker complexions.

Following Gebru’s announcement, she has received considerable support from individuals within the technology community.

Google declined to provide a statement for this report but referenced an internal email from Jeff Dean, Gebru’s supervisor. In that email, Dean stated: