During Jessie’s talk we discussed the ACLU’s coalition against facial recognition tech. Well, yesterday they revealed the results of a pretty incredible test they did. They took Amazon’s Facial Rekognition software, fed it 25,000 public arrest photos, then tried to match members of congress to those photos. The database falsely identified 28 members of congress as matching the arrest photos. 40% of those falsely identified were people of color, even though they only make up 20% of congress.
This reveals plainly what activists have been saying for some time – these technologies are deeply discriminatory and there is no way to use them safely, and there needs to be a moratorium on their use.