US lawmakers propose bill to tackle bias in tech firms’ algorithms
Image credit: Splosh | Dreamstime.com
US lawmakers proposed a bill on Wednesday 10 April that would require major tech companies to detect and remove any discriminatory biases in their computer models, indicating Washington’s aim to regulate Silicon Valley.
Called the Algorithmic Accountability Act of 2019, the bill would grant new power to the US Federal Trade Commission (FTC) and force companies to study if race, gender or other unconscious biases underpin their technology.
These rules would apply to companies with an annual revenue above $50m as well as data brokers and businesses with over a million consumers’ data.
“Computers are increasingly involved in the most important decisions affecting Americans’ lives – whether or not someone can buy a home, get a job or even go to jail,” Democratic senator Ron Wyden said when announcing the bill.
“But instead of eliminating bias, too often these algorithms depend on biased assumptions or data that can actually reinforce discrimination against women and people of colour,” he added.
The proposal comes as large tech companies grapple with accusations of racial, gender and political bias. It also underscores the growing bipartisan concern in Washington of large tech companies’ size and influence.
According to a Reuters report, Amazon Inc had scrapped an automated recruiting engine, which was found to be biased against women, and the US claims that Facebook has let advertisers discriminate by race in alleged violation of the Fair Housing Act.
Democratic senator Cory Booker and representative Yvette Clarke joined Wyden in introducing the bill, which could face an uphill battle in the Republican-controlled Senate.
It has also been supported by tech and civil rights groups such as Data on Black Lives, the Center on Privacy and Technology at Georgetown Law, and the National Hispanic Media Coalition.
“As long as humans are biased, algorithms will be biased too,” the Center on Privacy and Technology at Georgetown Law said in a statement. “This bill will force companies to reckon with that reality.”
Daniel Castro, vice president of the Information Technology & Innovation Foundation, added: “To hold algorithms to a higher standard than human decisions implies that automated decisions are inherently less trustworthy or more dangerous than human ones, which is not the case.”
In the midst of the proposed bill, Republican senators said Alphabet Inc’s Google, Facebook Inc and Twitter Inc discriminate against conservative viewpoints and suppress free speech, suggesting anti-trust action could be a solution.
Senator Ted Cruz, who chairs the Senate Judiciary subcommittee, said many Americans believe big tech firms are biased against conservatives. Cruz added that while no one wants “government speech police”, there are other remedies.
“Not only does big tech have the power to silence voices with which they disagree, but big tech has the power to collate a person’s feed in a way that agrees with their political agenda,” Cruz said in his opening statement at a US Senate Judiciary subcommittee hearing. “What makes the threat of political censorship so problematic is the lack of transparency.”
“If we have tech companies using the powers of monopoly to censor political speech, I think that raises real antitrust issues," Cruz added.
Facebook, Twitter and Google denied their platforms are politically biased and Democratic lawmakers said there was no evidence to back Republican bias claims, although Democrats have criticised the firms on other grounds.
Google had been disinvited to the subcommittee hearing over a dispute about whether it offered an executive senior enough to testify against bias allegations.
Cruz, however, said he plans a future hearing to address what he described as “Google’s censorship of free speech”.
Google said in a written statement submitted to the committee that it works to ensure “our products serve users of all viewpoints and remain politically neutral” but it acknowledged that “sometimes our content moderation systems do make mistakes”.
In September 2018, IBM launched a system which has detected unconscious bias in artificial intelligence algorithms.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.