UPDATED, May 23, 2019: Amazon shareholders have rejected two proposals that would have curbed the sales of its facial recognition software, Amazon Rekognition, according to Reuters and others. The vote breakdown is currently undisclosed, though an Amazon spokeswoman told Reuters the proposals failed by a wide margin.

The outcome is unsurprising following Amazon's pleas to the Securities Exchange Commission to stop the proposals from coming to a vote. Amazon's board had also recommended against the proposals.

The American Civil Liberties Union (ACLU), which strongly opposes the use of facial recognition technology, spoke out against the decision on Twitter. "This issue is not going away anytime soon," it wrote.

On the same day as the vote, the House Oversight Committee held a congressional hearing on the use of facial recognition software by law enforcement. Leaders came to a bipartisan agreement that the technology should be evaluated and regulated.

Dive Brief:

• Amazon shareholders have introduced two proposals to address human rights concerns over the company’s facial recognition software, known as Amazon Rekognition, reports the The New York Times.
• One proposal says the company should prohibit sales of Rekognition to government agencies unless the board decides a specific sale would not contribute to human rights violations. The second requests an independent report about whether the technology could "endanger, threaten or violate privacy or civil rights," and whether it could target people of color, immigrants or activists.
• The two resolutions are set for a vote at a shareholder meeting this Wednesday. Both are nonbinding, meaning Amazon would not have to take action even if they passed.

Dive Insight:

The resolutions signal growing concerns about facial recognition software as governments and private companies expand its use. Rekognition has been marketed to police agencies, and has been publicly tested in two pilots with the city of Orlando, FL. That’s prompted significant pushback, as activist groups and lawmakers worry that the software could be used to target certain groups, or may have inherent biases.

Studies — including one from Georgia Tech University and another from MIT — have shown that predictive technology can have trouble with darker skin tones. The MIT study said that "algorithmic justice necessitates a transformation in the development, deployment, oversight, and regulation of facial analysis technology." The American Civil Liberties Union (ACLU) uploaded 10,000 mugshots into Rekognition software last year and cross-referenced photos of members of Congress; the software spat out 28 false matches, disproportionately African-American.

San Francisco last week became the first city to ban city use of facial recognition technology, including by the police. Other cities, including Oakland, CA, are considering their own bans, and the U.S. House Committee on Oversight and Reform will hold a hearing on the technology’s potential impact on civil rights.

In a statement to The New York Times, Amazon said it has guidelines on how public safety agencies should use Rekognition, which include having a person review its results. "We have not seen law enforcement agencies use Amazon Rekognition to infringe on citizens’ civil liberties," the company said.

The votes on the resolutions will maintain pressure on Amazon as it tries to expand the market for Rekognition, including pitches to the Department of Homeland Security. In an open letter, the ACLU said lawmakers are "hearing the alarm bells" about what the group called "perhaps the most dangerous surveillance technology ever developed." Regardless of the outcome, the fact that Amazon is facing direct pressure on its technology will be cited as proof of the widespread concerns, which could offer more motivation to cities to step up their scrutiny.