The controversy over usage of facial recognition technology took center stage last week in Washington, D.C., as the House of Representatives’ Homeland Security Committee held a two-hour hearing, as opponents in the debate decry its racial bias and federal government’s quick rollout at U.S. airports without fully testing or acknowledging proven shortcomings.
In his opening remarks, Committee Chairman Bennie
G. Thompson (D-MS) outlined his concerns emerging from last July’s initial
hearing on the subject. “I am not wholly opposed to the use of facial
recognition technology, as I recognize that it can be valuable to homeland
security and serve as a facilitation tool for the Department’s varying missions,”
he stated. “But I remain deeply concerned about privacy, transparency, data
security, and the accuracy of this technology and want to ensure these concerns
are addressed before the Department deploys it further.”
Thompson cited a report from the National Institute of Standards and Technology (NIST) that confirmed age, gender, and racial bias in some facial recognition algorithms. Thompson noted NIST found that depending on the algorithm, African-American and Asian-American faces were misidentified 10 to 100 times more than white faces.
At the Feb. 6 hearing, officials of NIST, U.S.
Customs and Border Protection (CBP), and U.S. Department of Homeland Security
(DHS) were questioned by the committee, and generally defended facial
recognition as necessary to combat threats to national security and also help
airlines safeguard against a potential terrorist attack.
Thompson noted CBP touts that the match rate
for its facial recognition systems is over 98 percent, but it is his
understanding that NIST did not test CBP’s current algorithm for its December
2019 report. Nor CBP’s figure did not account for images of travelers who could
not be captured due a variety of factors such as lighting or skin tone – likely
making the actual match rate significantly lower.
“Leveraging CBP’s current authorities, we are
executing Congressional mandates to create and test an integrated biometric
entry-exit system using facial comparison technology,” testified John Wagner, deputy executive
assistant commissioner, CPB office of field operations. He added biometric
entry-exit is not a surveillance program and CBP does not use hidden cameras.
In response to a committee member’s question, Wagner
reported that 42.7 million people have already gone through the system, which
caught 252 imposters possessing traveling documents belonging to someone else.
He added he wasn’t aware of any false-positives.
Peter Mina, DHS’s deputy officer for programs
and compliance for Civil Rights and Civil Liberties, testified his office is
ensuring the use of facial recognition technology is consistent with civil
rights and civil liberties law and policy. Mina called for operators,
researchers, and civil rights policymakers to work together “to prevent
algorithms from leading to racial, gender, or other impermissible biases” in
the use of facial recognition technology.
Charles Romine, NIST’s director of the Information
Technology Laboratory, acknowledged in his testimony a general takeaway from his
agency’s biometric studies, which date back to the 1960s, a “significant
variance” between the performance facial recognition algorithms. “Consequently,
users, policy makers, and the public should not think of facial recognition as
either always accurate or always error prone.”
Consumer rights watchdog organizations remain
Jay Stanley, senior policy analyst for the American Civil Liberties Union (ACLU), analyzed government documents and statements, and concluded that DHS expansion of face recognition surveillance is already underway at airports despite insistence it’s a voluntary program.
He added the government policy threatens “a
dystopian future in which the technology is used throughout our public spaces
to scrutinize our identity, check us against watchlists, and record our
Congress needs to ban facial recognition, said Evan Greer, deputy director for Fight for the Future, in a press release prior to the hearing. “Biometric surveillance poses a unique threat to public safety and basic liberty,” said Greer, who believes there’s no safe way to use this technology. “Even if biased algorithms improve, face-scanning surveillance will automate and exacerbate existing forms of discrimination. It’s software for tyranny.”