What you’ll learn in this article:
- An opinion column published by The Oregonian from Washington, D.C.-based think tank Information Technology and Innovation Foundation says Portland should not ban facial recognition technology.
- The ITIF has board members who are government lobbyists from Amazon, Microsoft and other facial recognition developers that would benefit from proliferation of the technology.
- ITIF says its analysts operate separately from its board and development staff.
- To support its argument against a ban, the organization cites its own research, the interpretation of which is questionable.
- The article indicates many in government and corporate tech circles are watching Portland closely, and aim to influence the conversation around a potential ban.
UPDATE: This story was updated to include comments from ITIF on February 2, 2020.
A research analyst from a Washington tech and innovation think tank says Portland should test facial recognition tech rather than ban it. But why would a researcher in D.C. care so much about whether Portland outlaws some technology? It doesn’t take much surface-scratching to discover this guest columnist has some of the heavyweights of facial recognition behind her including Amazon and Microsoft, who also happen to be behemoths of Pacific Northwest tech.
“Rather than take a ‘ban first, ask questions later’ approach, Portland should undertake a series of small-scale pilots of [facial recognition] technology to evaluate its effectiveness and impact on privacy in various settings,” argues Ashley Johnson, a research analyst for the Information Technology and Innovation Foundation in a January 15 Oregonian opinion column.
“Rather than take a ‘ban first, ask questions later’ approach, Portland should undertake a series of small-scale pilots of [facial recognition] technology.”
– Ashley Johnson, ITIF
Portland has thoughtfully and methodically developed its approach to data privacy and tech policy since before adopting a privacy resolution in June. The city most recently unveiled draft rules that, if approved, would prohibit city government agencies including law enforcement from acquiring or using facial recognition. Also in the works: a ban on private use of the technology, such as in stores or apartment buildings. The city council plans to hold a second work session on January 28 addressing the potential ban, but it could be some time before the council votes on a final version.
Enter Information Technology and Innovation Foundation and its article in the state’s biggest newspaper. ITIF may not be well-known in Rip City, but the think tank is led by lobbyists working for some of tech’s best-known names – many of which would benefit from the proliferation of facial recognition.
Several ITIF board members are lobbyists for the biggest facial recognition players:
- Frederick Humphries, Jr, VP government affairs for Microsoft
- Cynthia Hogan, VP public policy for Apple
- Shannon Kellogg, VP of public policy for Amazon
- Jason Oxman, president and CEO of tech industry trade group International Technology Industry Council (members include Amazon, Google, Facebook and IBM)
Google lists ITIF among groups “that receive the most substantial contributions from Google’s U.S. Government Affairs and Public Policy team.” This 2017 Gizmodo article by Libby Watson provides a good roundup of the group’s funding, and its efforts to squelch regulations it views as anti-tech.
Daniel Castro, ITIF’s vice president said the group’s analyst team operates separately from its board as well as its development and fundraising staff. “I don’t even really interact with our board,” he said. “They don’t peer over our shoulders and get advanced copies of what we are going to say.”
The group will not name specific funders. Jackie Whisman, ITIF’s VP of development and outreach said in an email ITIF funding comes from “a diverse range of corporations, charitable foundations, government agencies, and individual contributors” and “corporate support comes from a diverse range of industries, including everything from advanced manufacturing to telecommunications, creative content, IT hardware, software and services, Internet, and life sciences.”
Winning Hearts and Minds
A quick search shows ITIF has been in the business of facial recognition influence peddling since at least 2018. The group has placed opinion articles in publications warning that facial recognition bans will hinder safety and stifle innovation. Its Vice President Daniel Castro gave testimony at a House Oversight Committee hearing on facial recognition technology the same day the Oregonian article ran.
The group joined a coalition of tech and security entities that sent an open letter to Congress in September. The letter’s message reflected the one in the Oregonian article, and noted that “Bans would keep this important tool out of the hands of law enforcement officers, making it harder for them to do their jobs efficiently, stay safe, and protect our communities.”
ITIF’s Oregonian article also argues that facial recognition is a helpful security tool. “One of the top benefits of facial recognition technology is improved public safety.”
Perhaps the most ardent supporter in local government of a ban has been Portland City Commissioner Jo Ann Hardesty (above). She told RedTail last year she believed outlawing government, police and private use of facial recognition would prevent the spread of tech that has a disparate impact on people of color. She also argued use of facial recognition technology by law enforcement could chip away at civil liberties and data privacy rights.
The Problem with Taking Research at Face Value
The ITIF column implies that city government’s evaluation of a facial recognition ban, particularly in the law enforcement context, is influenced by “inaccurate claims about the technology.” It mentions how some concern about facial recognition is based on an American Civil Liberties Union study which found that Amazon’s Rekognition software incorrectly identified 28 members of Congress as people who had been arrested for crimes.
ITIF argues, the ACLU’s “results have been shown to be spurious.” The article points to Amazon’s own defense against the ACLU’s findings. Amazon — a close affiliate of ITIF and a prominent facial recognition tech developer — said the ACLU’s test results were moot because the civil liberties defender did not apply the appropriate settings recommended by Amazon for public safety use of its system.
But, after damning the ACLU’s research, The ITIF article goes on to cite its own. And its interpretation of that research warrants some inspection. The article states:
“It is likely that the majority of Portlanders would not want the technology banned. In fact, a poll conducted by our partner organization found that fewer than one in five Americans agree with limitations on facial recognition technology that come at the expense of public safety, which clearly a ban would do.”
The “partner organization” that conducted the study is the Center for Data Innovation. Its top senior staff? Daniel Castro, ITIF’s own vice president, along with Eline Chivot, senior policy analyst for both organizations. In other words, for all intents and purposes, ITIF is citing its own study, one it conducted under the auspices of its closely-linked sibling.
The survey used in the study asked, “Agree or disagree? The government should strictly limit the use of facial recognition technology even if it comes at the expense of public safety.” Eighteen percent of respondents said they strongly or somewhat agree, while 55% said they somewhat or strongly disagree.
This 18% is what ITIF refers to in the article’s claim that “fewer than one in five Americans” is OK with limiting facial recognition even if it hinders public safety.
But 18+55=73. What happened to the remaining 27%?
The study results show that portion of respondents neither agreed nor disagreed with the statement. And the organization’s own chart highlighting results ignores this group entirely. But consider an interpretation combining those groups – the 18% who agree and the 27% who are unsure. That makes for a far more significant 45% who either agree or are unsure.
Policy through Honesty and Transparency
So, here we have instances of two organizations on opposite sides conducting their own research to support their opposing sides of a highly-contested debate that affects how governments craft policy for technologies that have life-altering impacts.
There is facial recognition research out there that is respected by both researchers in corporate tech and privacy and civil liberties groups; that’s the research provided by the National Institutes of Standards and Technology. In December, NIST finally published its long-awaited study of facial recognition algorithms and their demographic effects.
Scientific American’s coverage of the NIST study notes, “many of these algorithms were 10 to 100 times more likely to inaccurately identify a photograph of a black or East Asian face, compared with a white one. In searching a database to find a given face, most of them picked incorrect images among black women at significantly higher rates than they did among other demographics.”
No matter what side we’re on, we should all want reliable research when it comes to evaluating automated systems that could be used to decide if someone is arrested or if someone is allowed in a convenience store like the Jacksons in SE Portland (more on that here and in this video).
Like many others, this situation illustrates the need for media literacy. The NIST research itself states that “Reporting of demographic effects often has been incomplete in academic papers and in media coverage.” One glance at NIST’s most recent research illustrates the deep complexity of facial recognition, the issues it raises and the limits of splashy headlines, soundbytes and tweets.
We are immersed in an increasingly muddled information landscape where few people have the inclination or media literacy tools to inspect who’s behind the “expert” opinions offered in the local paper. But these opinions could sway attitudes about important government policy.
A discussion involving multiple voices and opinions is healthy. However, the conversation around Portland’s potential facial recognition ban is corrupted when tech giants seek to influence it by hiding behind the cloak of a seemingly impartial third party.