By: Matt Romano
Image Source: https://newsroom.cisco.com/feature-content?type=webcontent&articleId=1938827
Huge opportunities in the facial recognition industry has led tech giants like Amazon, Google, and Microsoft to develop their own facial recognition software. These companies are now selling this software to law enforcement and government agencies without any federal regulations on the technology.[1] Law enforcement all over the country are using this technology in criminal investigations by either running a photo through a database of faces or conducting a real- or near real-time analysis of video footage.[2] The concerns with these practices come from the fact that these facial recognition databases contain more than just mugshots. In 2016, a Georgetown Law study found that law enforcement’s use of facial recognition was affecting nearly half of all American adults.[3]
Earlier this year, the House of Representatives showed bipartisan concern for the technology’s potential violations of civil rights and privacy at a series of hearings, but no legislation is close to being passed at this time.[4] Without any federal regulations, some states have left the use of this technology by law enforcement vastly unregulated. In 2016, Georgetown Law found Florida was permitting law enforcement to review a database of millions of driver’s licenses without even requiring reasonable suspicion.[5] Moreover, eight thousand searches a month were being conducted within this database without the process being audited for misuse.[6] Along with concerns about privacy rights, several studies have shown evidence of racial and gender biases within the software algorithms.[7] These biases have drawn concerns that people of color are more likely to be misidentified in investigations, which could lead to wrongful convictions. As a response to these concerns many companies are working to rid their algorithms of these biases.[8]
While most states are trying to regulate the use of the technology,[9] Massachusetts, Michigan, and New York are considering temporary bans on the use of facial recognition technology by law enforcement until they can be certain that their citizens’ privacy can be protected.[10] Entire states banning the technology is obviously concerning to companies developing facial recognition software. They argue that, if regulated properly, the benefits far outweigh the risks.[11] Federal legislation would certainly help to reduce the risks to the public and give these companies consistency and clarity on what products they are permitted to sell. Amazon has grown tired of waiting for Congress draft laws themselves and has reportedly begun writing its own facial recognition laws to pitch to federal lawmakers.[12] It has not released any information on these laws yet, but they will likely mirror the guidelines for legislation it offered in a blog post from earlier this year.[13] Some of its recommendations in that blog post include requiring that the facial recognition system be ninety nine percent confident in a match for law enforcement investigations, providing written notice in areas where real-time facial recognition is in use, and developing standardized testing methodologies for measuring accuracy.[14] Although Amazon has been praised by some for taking an initiative to improve the industry, others have concerns about one of the wealthiest corporations in the world drafting laws that will govern its own product.[15] Whether you like Amazon drafting the laws or not, Congress must do something soon.
[1]See Jason Tashea, As Facial Recognition Software Becomes Ubiquitous, Some Governments Slam on the Brakes, Aba Journal (Sept. 24, 2019), http://www.abajournal.com/web/article/facial-recog-bans.
[2] See id.
[3] See Clare Garvie, Alvaro Bedoya & Jonathan Frankle, The Perpetual Line-Up, Geo. L. Ctr. on Privacy & Tech. (Oct. 18, 2016), https://www.perpetuallineup.org/ (finding that law enforcement facial recognition affects 117 million American adults).
[4] See Tashea, supra note 1.
[5] See Garvie, supra note 3.
[6]See id.
[7] See Drew Harwell, Amazon Facial Identification Software Used by Police Falls Short on Tests for Accuracy and Bias, Wash. Post (Jan. 25, 2019), https://www.washingtonpost.com/technology/2019/01/25/amazon-facial-identification-software-used-by-police-falls-short-tests-accuracy-bias-new-research-finds/#comments-wrapper; Steve Lohr, Facial Recognition is Accurate, if You’re a White Guy, N.Y. Times (Feb. 9, 2018), https://www.nytimes.com/2018/02/09/technology/facial-recognition-race-artificial-intelligence.html. But see Daniel Castro & Michael McLaughlin, Banning Police Use of Facial Recognition Would Undercut Public Safety, Info. Tech. & Innovation Found. (July 30, 2018),
https://itif.org/publications/2018/07/30/banning-police-use-facial-recognition-would-undercut-public-safety (providing evidence of flaws in a commonly referenced ACLU study on the bias of Amazon Rekognition).
[8] See Harwell, supra note 7 (confirming that Microsoft and IBM have improved their algorithms following the results of an independent study).
[9] See, e.g., Tashea, supra note 1 (mentioning that Vermont disallowed the search of it’s driver’s license databases by facial recognition).
[10] See id.
[11] See Daniel Castro, Are Governments Right to Ban Facial Recognition Technology?, Gov. Tech. (Apr./May 2019), https://www.govtech.com/products/Are-Governments-Right-to-Ban-Facial-Recognition-Technology.html (emphasizing facial recognition technology’s value in finding missing children, catching people with fake documents at airports, and combating human trafficking).
[12] See Jason Del Rey, Jeff Bezos says Amazon is Writing Its Own Facial Recognition Laws to Pitch to Lawmakers, Vox (Sept. 26, 2019), https://www.vox.com/recode/2019/9/25/20884427/jeff-bezos-amazon-facial-recognition-draft-legislation-regulation-rekognition.
[13] See id.
[14] See Michael Punke, Some Thoughts on Facial Recognition Legislation, AWS Machine Learning Blog (Feb. 7, 2019), https://aws.amazon.com/blogs/machine-learning/some-thoughts-on-facial-recognition-legislation/.
[15] See Del Rey, supra note 12 (“[W]e’ve seen this playbook before. Once companies realize that people are demanding strong privacy protections, they sweep in, pushing weak rules that won’t protect consumer privacy and rights.”).