Facial Recognition Know-how: Public Directors Should Face Up To The Moral Pitfalls

You may not know much about facial recognition software—but there’s a good chance facial recognition software knows something about you. At least one in two adults in the U.S. is registered in a law enforcement face recognition database, according to a 2016 report from Georgetown Law. The technology has quietly become increasingly common in our everyday lives, with myriad applications across retail, security, and even education. And while facial recognition and other biometric technologies stand to substantially improve efficiency within public services and effectiveness in law enforcement, they come with risks—risks that are today largely unchecked by regulations or oversight. The unfettered expansion of facial recognition technology has created a need for public administrators, in particular, to step up as responsible stewards of their constituents’ personal data and advocate on the side of privacy.

The primary use of facial recognition technology, so far, has been in the realm of security and public safety. The FBI uses closed-circuit television (CCTV), or video surveillance, with facial recognition capabilities at Border Patrol checkpoints, in airports, and other public places to spot face images of people involved in open investigations. But when it comes to ways the facial recognition technologies can be used, the sky’s the limit. In November, Delta opened the country’s first curb-to-gate biometric terminal at Atlanta’s Hartsfield Jackson International Airport, allowing customers to, for example, check in for a flight by facing a screen for two seconds. Facial recognition is being used in schools as a means to track who enters the building, and in retail stores to check customers against lists of recorded shoplifters.

The public stands to benefit from facial recognition technology that is thoughtfully and appropriately implemented. Governments can use technology to deliver public services more efficiently so that they require less human interaction and manual processing. Imagine a visit to the DMV to replace a lost driver’s license that requires you to bring no form of ID aside from your face, and takes much less time because of the significantly cut-down information processing time. Other public services could streamline their operations in similar ways whether they be county health clinics, homeless shelters, or City Hall.

These possible benefits are weighed down by a hefty load of ethical baggage. For one thing, the public’s facial images have already been gathered in databases by the FBI without their knowledge or consent, a revelation that provoked outrage when it was revealed at a March 2017 House Oversight Committee Hearing. Another alarming pattern displayed by this technology is that it’s too often unreliable, and is more likely to misidentify black people than white people. Use of these technologies could even be a violation of Constitutional rights because of the way they can limit  free speech by dissuading people to attend protests and speak openly in public for fear that they are under surveillance.

These potential benefits and significant drawbacks exist hand-in-hand, facing public administrators with an ethical conundrum: Should public administrators even consider leaning on this data for their public service and policy decisions, or would that in itself be an ethical violation?

It must come down to ensuring that facial recognition technologies meet a high ethical standard. A central piece of this standard should be affirmative consent: members of the public should be informed when their image is being captured, and they should have a clear opt-out option. Today, in the absence of any high-level regulation or oversight, public administrators must be the ones to step in and prevent any kind of experiment or pilot being conducted using facial recognition without public’s full knowledge. In addition, frank discussions should be initiated among top officials at local organizations like welfare agencies, economic development agencies, and local police departments, so that these leaders can be counted on to prioritize public consent and knowledge. At the same time, the federal government and all fifty states should pass laws that ensure oversight and accountability of anyone gathering and managing biometric data.

There’s no doubt that facial recognition and an ever-widening variety of biometric data technologies will become commonplace in both the public and private sectors in the coming years. There will be no returning to this nascent stage of implementation, which makes the time to address the many thorny ethical quandaries these technologies introduce. Public administrators are the ones who need to lead the charge in making sure that policies surrounding these technologies are forged in the interest of the public. By virtue of their role, public administrators are the bulwark between these technologies and the lives of the people they serve. This can’t be an area where they take that responsibility lightly.

Anirudh Ruhil is a Professor at Ohio University’s Voinovich School of Leadership and Public Affairs. He also directs the Voinovich Undergraduate Research Scholars program, a rewarding applied research program enrolling some of Ohio University’s most promising undergraduate students interested in leadership and public affairs. Ani also teaches quantitative research methods, policy analysis, and program evaluation for the Voinovich School’s Online Master of Public Administration academic programs. He serves as a quantitative research methodologist/data analyst for the School.