The development and expansion of facial recognition technology have accelerated over recent years, bringing with it significant benefits for a variety of organizations, ranging from law enforcement agencies to retail corporations.

From a security perspective, facial recognition software is being utilized in several ways. Recently, facial scanning technology was deployed at airports across the United States to help improve border patrol efforts and traveler safety. Facial recognition has also proven beneficial in the retail industry with technology designed to create better overall customer experiences.

Robin Lubbock/WBUR

On the surface, there’s seemingly little to argue against facial recognition technology. Software that simultaneously positively impacts aspects in both day-to-day life and the safety of the general public would expect to very popular.

However, as the use of facial recognition has grown across a variety of industries, questions have arisen surrounding the technology, calling into question its legality and potential infringements on personal privacy.

Let’s take a deeper look at some of the most common arguments against the expanded use of facial recognition technology in society.

Accuracy Concerns

The foundational benefits of facial recognition are found in the technology’s ability to identify people both quickly and accurately. Failure to accomplish both, especially accuracy, can have devastating effects on the people whom the technology is used.

Many detractors of facial recognition technology have pointed to recent research that shows an apparent discrepancy in facial recognition accuracy when it comes to both minorities and females.

In one study, researchers tested features of both Microsoft and IBM’s face-analysis services designed to identify the gender of people in photos. The findings discovered that the companies’ algorithms were nearly perfect in identifying the gender of men with lighter skin, but frequently misidentified images of women with dark skin.

Another review of face recognition software, conducted by Joy Buolamwini, an MIT Media Lab researcher and computer scientist, discovered the software misidentified the gender of dark-skinned females 35 percent of the time. In comparison, the error rate for light-skinned males was less than one percent.

The belief is that these errors are due to an underrepresentation of darker skin tones in the training data used to create the facial recognition software.

Given the current deployment of facial recognition software — for example, law enforcement agencies, air travel, etc. — misidentifications could have serious consequences, especially when it comes to public safety.

Until the accuracy issues are resolved and the risks of false positives are drastically lowered, many will continue to question the efficacy of facial recognition.

Need for Regulation

In the United States, every American is afforded the right to an expectation of privacy, thanks to the Fourth Amendment in the United States Constitution.

With that right in mind, however, facial recognition detractors are uneasy about the potential invasions of privacy as a result of the technology. Facial recognition software is already present in many aspects of day-to-day life, including brick-and-mortar stores where facial recognition technology is used to scan faces of shoppers to identify returning customers. Law enforcement teams have also deployed facial recognition technology in public and can identify someone by searching a biometric database that contains information on as many as one-third of Americans.

Perhaps the most compelling privacy concern with facial recognition is the lack of regulation on the technology. Currently, there aren’t laws against using facial recognition software nor any restrictions on how and where the technology can be used.

Technology providers themselves have urged policymakers to study and monitor the use of the facial recognition industry. In July 2018, Microsoft compared the industry to that of cars and medicine which require government regulation.

“We live in a nation of laws, and the government needs to play an important role in regulating facial recognition technology,” Bradford Smith, the company’s president, wrote in a blog post. He added: “A world with vigorous regulation of products that are useful but potentially troubling is better than a world devoid of legal standards.”   

Proponents of facial recognition see the technology as an incredible tool to improve law enforcement and security efforts, identifying, and potentially, locating criminals with unmatched speed and accuracy. The lack of regulation, on the other hand, has its opponents envisioning an eventual evolution into a mass surveillance society where the day-to-day constitutional privacy rights of Americans could be potentially violated without any explicit consent given.

Room for Improvement

Regardless of the valuable benefits that facial recognition has shown, it is not perfect. It is not a technology devoid of making mistakes, and as a result, there are plenty of people who see its expansion, at the very least, controversial.

From removing potential bias in the software to ensuring the facial recognition industry operates by specific rules and regulations to protect the privacy of those it surveils, there is indeed room for improvement when it comes to the technology.

And until these problems are resolved, facing these issues — and facial recognition detractors — will be a consistent challenge for technology companies.