Amazon's facial recognition moratorium has major loopholes

Zack Whittaker

In a surprise blog post, Amazon said it will put the brakes on providing its facial recognition technology to police for one year, but refuses to say if the move applies to federal law enforcement agencies.

The moratorium comes two days after IBM said in a letter it was leaving the facial recognition market altogether. Arvind Krishna, IBM's chief executive, cited a "pursuit of justice and racial equity" in light of the recent protests sparked by the killing of George Floyd by a white police officer in Minneapolis last month.

Amazon's statement — just 102 words in length — did not say why it was putting the moratorium in place, but noted that Congress "appears ready" to work on stronger regulations governing the use of facial recognition — again without providing any details. It's likely in response to the Justice in Policing Act, a bill that would, if passed, restrict how police can use facial recognition technology.

"We hope this one-year moratorium might give Congress enough time to implement appropriate rules, and we stand ready to help if requested," said Amazon in the unbylined blog post.

But the statement did not say if the moratorium would apply to the federal government, the source of most of the criticism against Amazon's facial recognition technology. Amazon also did not say in the statement what action it would take after the yearlong moratorium expires.

Amazon is known to have pitched its facial recognition technology, Rekognition, to federal agencies, like Immigration and Customs Enforcement. Last year, Amazon's cloud chief Andy Jassy said in an interview the company would provide Rekognition to "any" government department.

Amazon spokesperson Kristin Brown declined to comment further or say if the moratorium applies to federal law enforcement.

There are dozens of companies providing facial recognition technology to police, but Amazon is by far the biggest. Amazon has come under the most scrutiny after its Rekognition face-scanning technology showed bias against people of color.

In 2018, the ACLU found that Rekognition falsely matched 28 members of Congress as criminals in a mugshot database. Amazon criticized the results, claiming the ACLU had lowered the facial recognition system's confidence threshold. But a year later, the ACLU of Massachusetts found that Rekognition had falsely matched 27 New England professional athletes against a mugshot database. Both tests disproportionately mismatched Black people, the ACLU found.

Investors brought a proposal to Amazon's annual shareholder meeting almost exactly a year ago that would have forcibly banned Amazon from selling its facial recognition technology to the government or law enforcement. Amazon defeated the vote with a wide margin.

The ACLU acknowledged Amazon's move to pause sales of Rekognition, which it called a "threat to our civil rights and liberties," but called on the company and other firms to do more.

Amazon shareholders want the company to stop selling facial recognition to law enforcement


More From

  • Apple allocates its first $400M from $2.5B commitment to address California's housing crisis

    Apple announced this morning it's allocating more than $400 million toward affordable housing projects and other homeowner assistance programs in California, as a part of its earlier multi-year pledge of $2.5 billion to address the state's housing crisis and homelessness issues. The funding is expected to support thousands of Californians with first-time homebuyer assistance or new, affordable housing units, Apple says. Projects launching in 2020 include 250 new units of affordable housing across the Bay Area -- the first affordable housing developments funded in a private-public partnership with Housing Trust Silicon Valley.

  • India seeks new regulator for non-personal data

    India should set up a data regulator to oversee how companies collect, process, store, monetize and even destroy non-personal data (or data that has been anonymized), a panel tasked by New Delhi has recommended in a draft report. The eight-person panel said that companies such as Google, Facebook, Amazon, and Uber have benefited from a combination of “first mover advantage,” “sizable network effect” and “enormous data” that they have collected over the years. This dominance has “left many new entrants and startups being squeezed and faced with significant entry barriers,” said the draft report, which has been made available to industry players for consultation before it is submitted to the nation's IT ministry next month.

  • Equity Monday: India's digital economy attracts ample attention, three funding rounds, and earnings season

    This is Equity Monday, our week-starting primer in which we go over the latest news, dig into the week ahead, talk about some neat funding rounds and dive into the latest big news from the startup world. Qualcomm pours capital into India's Reliance Jio, which has now raised $15.7 billion in the past three months. Google intends to invest $10 billion into India over the next five to seven years, looking to put capital to work in through a mix of "a mix of equity investments, partnerships, and operational, infrastructure and ecosystem investments" via a new "Google for India Digitization Fund."

  • SiriusXM buys Stitcher for $325 million, steps up its march into podcasts

    Less than a month after picking up Simplecast for its podcast distribution and analytics tools, SiriusXM today is announcing an even bigger acquisition to raise its game in the realm of streamed spoken-word content. The satellite radio company said it has reached a deal to acquire Stitcher from E.W. Scripps for $325 million, a return of more than double Scripps' investment in podcasting, the companies reported this morning. Stitcher is a podcast pioneer that provides a popular one-stop platform to create, monetize (via advertising) and distribute podcasts to listen to via its app and on multiple platforms, and that is precisely what SiriusXM would like to capitalise on.