Researchers in Support of Amazon Employees: Amazon should not sell Face Rekognition Technologies to Police, Government or Intermediaries.

An Open Letter to:

Jeffrey Bezos, Founder and CEO of Amazon.com Inc.
Hassan Sawaf, Director of Artificial Intelligence, Amazon Web Services (AWS)
Dr. Matt Wood, General Manager of Artificial Intelligence, Amazon Web Services (AWS)

As academics who study, teach about, and develop information technologies, we write in solidarity with Amazon workers and shareholders, as well as a coalition of over 70 civil and human rights groups, who are urging Amazon to cease sale of its Rekognition facial recognition technology to police departments and government agencies. We also support the workers’ demand to cease providing AWS cloud services to intermediary companies such as Palantir that develop AI technologies for police departments and militaries, and their refusal to provide services to the U.S. Immigration and Customs Enforcement (ICE) service, and refusal to design tools that facilitate human rights violations, such as the U.S. Department of Homeland Security’s Homeland Advanced Recognition Technology (HART) database of biometric data.

As two of the authors of this letter have forcefully argued, face recognition technology represents an unprecedented threat to privacy and civil liberties that even the best traditional forms of regulation will almost certainly fail to mitigate. Amazon has an obligation to consider the privacy harms, and threats to civil and human rights, from the infrastructure of cameras, face databases, and biometric data its products are creating.

Recent studies of facial recognition systems deployed in the United Kingdom have shown unacceptably low levels of accuracy. However, even a highly accurate, ubiquitous identification and tracking system itself poses a grave threat to society—a matter on which there is bipartisan agreement even in this sharply divided political climate. Brian Brakeen, CEO of the facial recognition company Kairos, has acknowledged that their own technology presents such serious privacy concerns that they will not sell it to governments, while the City of Orlando, Florida has dropped its use of Rekognition following the expression of privacy concerns.

Amazon’s insistence that there has been “no reported law enforcement abuse of Amazon Rekognition” is a woefully insufficient response. Just as Amazon sees no evidence of abuse of Rekognition by law enforcement, we see no evidence of a sufficient reporting system for such abuse. Nor does there appear to be any established system for documenting, let alone preventing, the use of Rekognition for illegal discrimination, or in violation of human, civil or privacy rights. Indeed, this technology would amplify existing forms of implicit bias against vulnerable minorities, and enable even more powerful forms of explicit discrimination and oppression.

It is incumbent upon powerful companies like Amazon to sharply restrict the sale of products such as Rekognition, while they work to ensure greater transparency and accountability in the development and deployment their technologies.

We thus call upon Amazon to:

• Stop supplying government and law enforcement agencies with facial recognition technology.
• Cancel its cloud services contracts with Palantir and other intermediary companies that provide data analytics and AI services to police or militaries.
• Establish ethical guidelines that prohibit the weaponization and militarization of this and other technologies that threaten privacy, civil and human rights, and establish transparency and accountability mechanisms to ensure those guidelines are implemented.

__________________________________________________________

Add Your Signature to this Letter:

[pdb_signup]

[pdb_list]