‘Facial Recognition technology is faulty, racist, biased, abusive to civil rights; act now to restrict misuse’ say experts to House Oversight and Reform Committee

Last week, the US House Oversight and Reform Committee held its first hearing on examining the use of ‘Facial Recognition Technology’. The hearing was an incisive discussion on the use of facial recognition by government and commercial entities, flaws in the technology, lack of regulation and its impact on citizen’s civil rights and liberties.

The chairman of the committee, U.S. Representative of Maryland, Elijah Cummings said that one of the goals of this hearing about Facial Recognition Technology is to “protect against its abuse”.

At the hearing, Joy Buolamwini, founder of Algorithmic Justice League highlighted one of the major pressing points for the failure of this technology as ‘misidentification’, that can lead to false arrests and accusations, a risk especially for marginalized communities. On one of her studies at MIT, on facial recognition systems, it was found that for the task of guessing a gender of a face, IBM, Microsoft and Amazon had error rates which rose to over 30% for darker skin and women. On evaluating benchmark datasets from organizations like NIST (National Institute for Standards and Technology), a striking imbalance was found. The dataset contained 75 percent male and 80 percent lighter skin data, which she addressed as “pale male datasets”. She added that our faces may well be the final frontier of privacy and Congress must act now to uphold American freedom and rights at minimum.

Professor of Law at the University of the District of Columbia, Andrew G. Ferguson agreed with Buolamwini stating, “Congress must act now”, to prohibit facial recognition until Congress establishes clear rules. “The fourth amendment won’t save us. The Supreme Court is trying to make amendments but it’s not fast enough. Only legislation can react in real time to real time threats.”

Another strong concern raised at the hearing was the use of facial recognition technology in law enforcement. Neema Singh Guliani, Senior Legislative Counsel of American Civil Liberties Union, said law enforcement across the country, including the FBI, is using face recognition in an unrestrained manner. This growing use of facial recognition is being done “without legislative approval, absent safeguards, and in most cases, in secret.” She also added that the U.S. reportedly has over 50 million surveillance cameras, this combined with face recognition threatens to create a near constant surveillance state.

An important addition to regulating facial recognition technology was to include all kinds of biometric surveillance under the ambit of surveillance technology. This includes voice recognition and gait recognition, which is also being used actively by private companies like Tesla. This surveillance should not only include legislation, but also real enforcement so “when your data is misused you have actually an opportunity, to go to court and get some accountability”, Guliani added. She also urged the committee to investigate how FBI and other federal agencies are using this technology, whose accuracy has not been tested and how the agency is complying with the Constitution by “reportedly piloting Amazon's face recognition product”.

Like FBI and other government agencies, even companies like Amazon and Facebook were heavily criticized by members of the committee for misusing the technology. It was notified that these companies look for ways to develop this technology and market facial recognition.

On the same day of this hearing, came the news that Amazon shareholders rejected the proposal on ban of selling its facial recognition tech to governments. This year in January, activist shareholders had proposed a resolution to limit the sale of Amazon’s facial recognition tech called Rekognition to law enforcement and government agencies. This technology is regarded as an enabler of racial discrimination of minorities as it was found to be biased and inaccurate.

Jim Jordan, U.S. Representative of Ohio, raised a concern as to how, “Some unelected person at the FBI talks to some unelected person at the state level, and they say go ahead,” without giving “any notification to individuals or elected representatives that their images will be used by the FBI.”

Using face recognition in such casual manner poses “a unique threat to our civil rights and liberties”, noted Clare Garvie, Senior Associate of Georgetown University Law Center and Center on Privacy & Technology. Studies continue to show that the accuracy of face recognition varies on the race of the person being searched. This technology “makes mistakes and risks making more mistakes and more misidentifications of African Americans”. She asserted that “face recognition is too powerful, too pervasive, too susceptible to abuse, to continue being unchecked.”

A general agreement by all the members was that a federal legislation is necessary, in order to prevent a confusing and potentially contradictory patchwork, of regulation of government use of facial recognition technology. Another point of discussion was how great facial recognition could work, if implemented in a ‘real’ world. It can help the surveillance and healthcare sector in a huge way, if its challenges are addressed correctly. Dr. Cedric Alexander, former President of National Organization of Black Law Enforcement Executives, was more cautious of banning the technology. He was of the opinion that this technology can be used by police in an effective way, if trained properly.

Last week, San Francisco became the first U.S. city to pass an ordinance barring police and other government agencies from using facial recognition technology. This decision has attracted attention across the country and could be followed by other local governments.

A council member Gomez made the committee’s stand clear that they, “are not anti-technology or anti-innovation, but we have to be very aware that we're not stumbling into the future blind.”

Cummings concluded the hearing by thanking the witnesses stating, “I've been here for

now 23 years, it's one of the best hearings I've seen really. You all were very thorough and very very detailed, without objection.”

The second hearing is scheduled on June 4th, and will have law enforcement witnesses.

For more details, head over to the full Hearing on Facial Recognition Technology by the House Oversight and Reform Committee.

Over 30 AI experts join shareholders in calling on Amazon to stop selling Rekognition, its facial recognition tech, for government surveillance

Oakland Privacy Advisory Commission lay out privacy principles for Oaklanders and propose ban on facial recognition

Is China’s facial recognition powered airport kiosks an attempt to invade privacy via an easy flight experience?