ACLU Comment on New Amazon Statement Responding to Face Recognition Technology Test

SAN FRANCISCO — Amazon today issued an additional statement in response to the American Civil Liberties Union Foundation of Northern California test of Rekognition, the company’s face recognition technology. The test revealed that Rekognition falsely matched 28 current members of Congress with images in an arrest photo database. Congressional members of color were disproportionately identified incorrectly, including six members of the Congressional Black Caucus.

Jacob Snow, technology and civil liberties attorney at the ACLU Foundation of Northern California, responded with the following comments:

“In its five stages of grief over its dangerous face surveillance product, Amazon is clearly stuck at denial. In a matter of 48 hours, Amazon has gone from its own system default of an 80 percent match rate to saying yesterday it should be 95 percent, and then saying today it should be 99 percent. At no time has Amazon taken any responsibility for the very grave impact that their face surveillance product has on real people.

“Instead, Amazon is grasping at straws in an attempt to distract from critical civil rights issues. Amazon should take steps to fix the damage its ill-advised face surveillance product may have already caused and to prevent further harm. Amazon should respond to members of Congress. It should disclose every government agency that has already purchased this technology. And it should heed the calls of organizations and its own customers, employees, and shareholders and stop selling face surveillance to the government. The fact that Amazon has refused to address the very real threats its technology poses, let alone take these necessary actions, is further evidence of its disappointing state of denial – and the need for Congress to quickly step in with a moratorium.”

Snow’s comments in response to Amazon’s initial statement can be found here: https://www.aclu.org/news/members-congress-demand-meeting-amazon-ceo-fol….

Below is a timeline of developments to date:

• February 2017 — The Sheriff’s office in Washington County Oregon (correctly) warned Amazon their work together could raise civil liberties concerns, stating that the “ACLU might consider this the government getting in bed with big data.”
• June 2017 – Amazon and Washington County write a blog post together instructing people on how to use Rekognition to find people in arrest photos, just as the ACLU did in its test. That post sets the confidence threshold at 85% (not 95% and certainly not 99%).
• May 2018 — ACLU releases the results of an investigation showing that Amazon was aggressively marketing face recognition to law enforcement as a tool of mass surveillance.
• June 18, 2018 — The ACLU, together with local community partners, deliver a petition with over 150,000 signatures and coalition letter signed by nearly 70 organizations to Amazon headquarters in downtown Seattle. That same day, 19 investor groups echo the ACLU’s demand in a shareholder letter to Mr. Bezos. In their letter, shareholders warn the technology “may not only pose a privacy threat to customers and other stakeholders across the country, but may also raise substantial risks for our Company, negatively impacting our company’s stock valuation and increasing financial risk for shareholders.”
• June 21, 2018 —Amazon employees send a powerful letter to Jeff Bezos calling on him to stop selling Rekognition to law enforcement, stating that they “refuse to contribute to tools that violate human rights” and that Amazon “should not be in the business of supporting those who monitor and oppress marginalized populations.”
• July 26, 2018 — ACLU releases the results of its independently verified test showing that, using the default match settings, 28 members of congress incorrectly matched arrest photos, with members of congress of color disproportionately being falsely identified.
• July 26, 2018 — Amazon states that it guides law enforcement customers to set a threshold of 95% for face recognition. Amazon also notes that, if its face recognition product is used with the default settings, it won’t “identify[] individuals with a reasonable level of certainty.”
• July 27, 2018 — Amazon writes that even 95% is an unacceptably low threshold, and states that 99% is the appropriate threshold for law enforcement.