"It's not clear exactly who decided that Gebru's paper had to be quashed or for what reason. Nor is it clear why her resistance—predictable as it was—prompted a snap decision to eject her, despite the clear risk of public fallout. Other researchers at Google say it isn't unusual for publications about AI to trigger internal corporate sensitivities before public release, but that researchers can usually work through managers' objections. Gebru, with her track record of rattling management about Google's diversity and AI ethics problems, got little such opportunity. One reason managers were not more open in explaining their feedback to Gebru, according to Google, was that they feared she would spread it around inside the company. Those fears may have been compounded when Gebru took to an internal listserv to criticize Google for “silencing marginalized voices,” even as she offered to kill her own paper in exchange for greater transparency."