Google fired a top AI researcher on Friday, the second high-level departure to spotlight both the management of sensitive artificial intelligence issues and broader internal dissatisfaction with race and diversity.
The internet group said Margaret Mitchell, who had co-headed the AI ethics group, had been fired for “multiple violations of our code of conduct” and security policy. These include “exfiltrating confidential company sensitive documents and other employees’ private data,” it said.
The move comes two months after the departure of Timnit Gebru, the other head of the ethics group, who was created to lead research into new techniques to ensure the company’s AI technology was used in a fair and unbiased manner. .
Gebru said she was fired after the company blocked the publication of a research paper she co-authored on the potential bias in AI systems for languages like those used by Google, while the company said she chose to leave.
Gebru’s departure also highlighted Google’s continued dissatisfaction with its approach to racial and gender diversity issues.
In a letter published at the time of Gebru’s departure from the company and later made public, Mitchell claimed the dispute stemmed from assault on her fellow investigator.
“The shooting seems to have been fueled by the same underpinnings of racism and sexism that our AI systems absorb when in the wrong hands,” she said.
She added that Gebru “had refused to submit herself to a system that requires her to belittle her integrity as a researcher and degrade herself among her fellow researchers.”
Sundar Pichai, Google’s CEO, apologized internally for Gebru’s loss and promised an internal review.
“We have to take responsibility for the fact that a prominent black female leader with enormous talent left Google unfortunately,” he wrote to staff in an email. “This loss has had a ripple effect in some of our least represented communities, who have seen themselves and some of their experiences reflected in Dr. Gebru’s.”