(Reuters) – US lawmakers proposed a bill on Wednesday that would require large tech companies to discover and remove discriminatory phenomena embedded in their computer models and underline Washington's growing interest in regulating Silicon Valley.
Bill justified The law on algorithmic accountability from 2019, would give new power to the US Federal Trade Commission and force companies to study whether race, gender or other prejudices underlie their technology. The rules would apply to companies with annual revenues in excess of $ 50 million, as well as computer hardware companies and companies with over one million consumer data.
"Computers are increasingly involved in the key decisions affecting Americans' lives ̵
The press release, for example, was cited as a Reuters report that Amazon.com Inc. had scrapped an automated recruiting engine that it turned out to be biased against women, and the United States accuses Facebook Inc. of allowing advertisers to discriminate through race in alleged violation of the Fair Housing Act.  Sen. Cory Booker and Rep. Yvette Clarke, both Democrats, joined the Wyden for the introduction of the bill, which could face an upward blow in the Republican-controlled Senate.
"To keep algorithms To a higher standard than human decisions, automated decisions are in themselves less credible or more dangerous than human, which is not the case, says Daniel Castro, vice president of the Information and Innovation Foundation, a Washington-based non-profit organization that includes industry representatives on the board.
"This would only serve to stigmatize and discourage AI use, which can reduce its beneficial social and economic impact," Castro says.
The Internet Association, which counts with Amazon, Facebook, Alphabet Inc.'s Google, and other top tech companies as members, had no immediate comment.