Amazon shareholders “will vote against facial recognition”
Shareholders in Amazon will vote next week to ban the company from using facial recognition techniques, in a symbolic move against the retail giant’s development of deep learning tools.
While the vote is not binding and may not be accepted by the company, shareholder groups are hoping that it will send a message to Amazon that techniques such as facial recognition should not be used until they can be proven to be safe.
OpenMIC, the non-profit group leading the vote, has demanded that Amazon’s board stop pursuing these techniques until an independent arbitrator can determine that they do not “cause or contribute to actual or potential violations of human rights.”
Amazon’s facial recognition tech is touted as being able to carry out tasks such as identifying individuals in crowds and, for example, matching them against entries on watchlists.
But potential human rights violations have been of deep concern to US civil liberties groups. The American Civil Liberties Union (ACLU) tested Amazon’s facial recognition software and found that it falsely identified more than two dozen members of Congress to suspected criminals.
Human rights organisations are also concerned that the technology could be used to clamp down on legitimate protest.
Amazon, along with Microsoft, have called for oversight to be imposed not on the companies themselves, but only on the technology. In a previous joint statement they said: “It’s critical that any legislation protects civil rights while also allowing for continued innovation and practical application of the technology.”