In “Facial recognition faces a reckoning,” BBC’s Tech Tent wonders whether it is enough for tech giants to simply stop selling their AI systems to law enforcement. Following the death of George Floyd, IBM decided to stop selling its facial technology due to “inherent bias”. Both Amazon and Microsoft have halted the sale of their respective products to law enforcement. The two tech giants have called for Congress to pass legislation governing the use of facial recognition software.
Facial Recognition Built on Bias?
Amazon’s Facial Rekognition has been under scrutiny since it came out. Tech Tent notes how a former Amazon employee, Anima Anandkumar, stated that Facial Rekognition was launched before it was ready. “Researchers like me wanted to take a slower approach,” Anandkumar, who is now a professor of computing at California Institute of Technology said.
Facial recognition technology uses datasets to train artificial intelligence and machine learning systems to be able to take decisions on their own. Anima argues that facial recognition tech relies heavily on imbalanced datasets where black bodies – men and women – are unfairly represented.
Facial Technology is Big Business
For analyst Stephanie Hare, the problem goes beyond police departments around the world. She notes that facial technology software is big business for both private and public sectors. For her, governments would have to address other ethical issues such as the accessibility of facial technology as well as options to opt out.
Should facial recognition technology be banned?
Mobile Techer 