Racial bias in facial recognition algorithms
Amnesty International CanadaJun 08, 2023
The US Federal Trade Commission (FTC) on Tuesday prohibited Rite Aid from using artificial intelligence-based facial recognition technology for five years in an agreement that settles charges that the bankrupt pharmacy chain's AI technology harmed consumers.
Rite Aid reportedly used facial recognition technology for surveillance purposes, but the FTC claims that the system misidentified thousands of customers as potential shoplifters, disproportionately affecting non-White shoppers.
Rite Aid’s settlement is yet another example of how AI is riddled with structural racism, just like society at large is. Rite Aid’s facial recognition software produced thousands of false positives, mostly by misidentifying people of color, putting marginalized groups at further risk. This issue isn't exclusive to the pharmacy, as countless studies show that AI has a racial bias and is discriminatory. AI needs an overhaul to prevent such practices before it can be a prominent part of society.
While AI may be a novel development that brings new issues to the forefront, it can't escape America’s favorite pastime: complaining about racism. Despite racism being legally and morally weeded out of the West, the mainstream media loves to play up any instance of perceived bias. The left has been brainwashed into thinking that group differences don’t exist and that any unequal outcomes are purely caused by social factors. AI doesn’t buy this woke falsehood yet, which is why it's falsely dubbed “racist.”