Uber faces new racial discrimination claims from drivers who claim they were wrongly fired due to malfunctioning facial recognition technology.
The complaints emerged after Uber introduced an automated system to verify the identity of drivers operating its services in April last year. Every time a driver checks in for work, they have to take a selfie photo which is then compared, using an automated system, to that of their Uber account profile.
Pa Edrissa Manjang, who worked for Uber Eats take-out courier in London, filed a complaint with the employment tribunal alleging that her account had been illegally deactivated. He says automated facial verification software mistakenly decided that his selfie photos were someone else’s on several occasions.
Manjang, whose case is supported by the Drivers & Couriers Union App, was fired on May 1 of this year for “persistent inconsistencies” between the photos he took to register for a job and that in his profile. Uber professional. When he suggested that Uber ask a human to review the photos, he was told that after “careful consideration” his account was being deactivated. No further information was provided on the nature of the examination of his case.
Manjang claims that Uber did not investigate the potential discriminatory effect of the automated software, to tell him about the problem or to allow a human to review the photos he had recorded with.
Another person who filed a complaint with the court is Imran Javaid Raja, who was fired in October 2020 after the same checks led to his license being revoked by Transport for London. He was reappointed the following month, with Uber admitting they had made a mistake. However, Raja was unable to work until the following January when his TfL license was reinstated and he says he never received compensation for his period of unemployment.
“Uber should not be allowed to use facial recognition software in the UK against a vulnerable workforce already at risk of exploitation and human rights violations,” the union said.
He wrote to Microsoft about the use of its technology at Uber and the company said those deploying its facial recognition software should incorporate “meaningful human examination to detect and resolve instances of misidentification or misidentification. other failures “.
In 2019, Microsoft, which manufactures the software, conceded that facial recognition software didn’t work as well for people of color and might not recognize them.
The latest complaints against the use of the software emerged after the Self-Employed Workers Union of Great Britain said the technology had wrongly led at least 36 drivers to have their registration with Uber terminated since the start of the campaign. pandemic. He asks Uber to remove the “racist algorithm” and reinstate the dismissed drivers.
The question is particularly relevant in London, where nine in ten private drivers are black or black British, Asian British or Asian or mixed race, according to one. TfL survey.
The IWGB is backing a lawsuit against another anonymous Uber driver who lost his job when automated facial analysis software failed to recognize him.
IWGB members working for Uber staged a 24-hour strike on Wednesday and protested outside the company’s London headquarters over the issue of facial recognition and wages.
Uber previously said there are two manual manual reviews before any decision to remove a driver is made and that the system is “fair and important to the safety of our platform.” He also said anyone removed from the platform could appeal the decision. Uber has been contacted for comment.
Studies of several facial recognition software packages showed higher error rates when recognizing people with darker skin than among people with lighter skin, although Microsoft and others improved their performance.
Microsoft declined to comment on an ongoing court case.