In January, Detroit police came to the home of a Black man named Robert Williams and arrested him on his front lawn in front of his wife and two young daughters, then locked him in a filthy cell for 30 hours, according to a lawsuit that the ACLU filed this week. It turned out to be a case of mistaken identity, the suit alleges, based on a faulty match in a face recognition system.
The case is just one maddening example of how technologies developed by a mostly white and Asian workforce in Silicon Valley and Seattle can end up disproportionately harming Black people. The tech industry has known for years that face recognition systems are far more likely to misidentify people with darker skin. And yet only last week, at the height of the nationwide protests against police brutality that followed the police killing of George Floyd, did IBM, Microsoft, and Amazon announce that they would halt development of those systems for law enforcement, and only temporarily in Amazon’s case. Even then, other tech firms quickly jostled to claim their market share.
Momentum is a blog that captures and reflects the moment we find ourselves in, one where rampant anti-Black racism is…
The protests, and renewed outcry from both Big Tech’s critics and its own employees, have focused the attention of Silicon Valley leaders on the industry’s racial inequities, and its products’ discriminatory impacts, as never before. And yet the partial pullback from face recognition stands as one of the few tangible changes to emerge from this reckoning so far. As the Los Angeles Times reported in depth this week, the industry has made scant progress in Black and Latino representation, and Black workers still routinely face racism in the workplace. Some of its apparently well-intended efforts have only served to underscore the problem, as when Snapchat released a filter in honor of Juneteenth that asked users to smile in order to break the chains of bondage.
“You can hire employees all day, and if…