Disadvantages of The Application of Facial Recognition Technology in Law Enforcement

Introduction: With the advance of technology, artificial intelligence has become an important part of human life. Facial recognition technology, which is one of the artificial intelligence, has become gradually fashionable in human life. Background: Admittedly, facial recognition technology brings a lot of convenience to human life, so some people asked whether facial recognition technology can be applied to law enforcement. After doing some research, the answer is no. Thesis statement: There are three disadvantages of facial recognition technology, which makes it unable to be applied in law enforcement: low accuracy, racial discrimination, and invasion of privacy. This paper explores the shortcomings of facial recognition technology in law enforcement based on examples and statistics. It is hoped that this paper will bring new ideas and enlighten the development of facial recognition technology.

Introduction

In 2018, Jacky Cheung (a famous Chinese star) held a concert in Nanchang Province. At that time, due to the heavy rain and cooling in Nanchang, the weather was not very good, but it did not dampen the enthusiasm of fans, even an escaped criminal secretly came to the concert. Although the fugitive was wearing a decent suit, he was taken away by several police officers a moment later. This is because policemen used on-site facial recognition technology to identify the audience. When discovering his identity, through the security system positioning, policemen could successfully catch the fugitive. This is an example that shows the application of facial recognition technology in law enforcement.

You might be wondering how does facial recognition work? Facial recognition technology is usually photographed by a digital camera and then analyzes human facial features. It scans the entire face first, analyses the face structure next, and shows the distribution of the five sense organs of the tested face. These unique data are kept in reserve in a database for later comparison. In the previously stated example, policemen used facial recognition technology to analysis the facial features of this criminal and compared them with the information in the database. Hence, the outlaw was incarcerated and he was kept in a prison.

However, whether facial recognition technology should be used in law enforcement is a controversial topic. Some people agree that law enforcement should use facial recognition technology, their reasons are that facial recognition can find criminals accurately, increase the detection rate and keep society safe. However, others hold opposite views. Many people do not think that it should be used as a tool to help law enforcement work. Also, they do not think it can help law enforcement maintain social order and ensure people's safety. There are three reasons why facial recognition technology should not be used in law enforcement: poor reliability, racial bias, and intrusion of privacy.

Analysing Reasons

Topic sentence: First of all, one of the disadvantages of facial recognition technology is that its accuracy is not as high as people expect. Evidence & citing: The facial recognition technology requires to build a face model, which is a database of facial characteristics that are collected. When law enforcement uses facial recognition technology to identify criminals, they need to compare the information in the database with the images collected during the criminal investigation. If there is no same face in the face model, the system may provide the most likely candidate photographs. This requires manual inspection by the police. Moreover, if the suspect's information is not in the database, then facial recognition technology is useless. Therefore, the accuracy of facial recognition technology is not 100%.

Furthermore, the results of facial recognition technology may trigger the wrong 'match' or the wrong 'reject' for various reasons. Some reasons affect the accuracy of the results: poor light, glare from glasses, the change of facial hair or hairstyle, and the angle of the photographs. Even in a controlled situation, the results are still not completely accurate. When taking personal photos at random on the street, the error is greater, because a surveillance camera is often unable to take front-facing photos. Ideally, the person who is photographed should be under strict control. For example, the subject should look directly at the camera and fill in the photo area, so that an automatic system can reliably identify the person in the photo and even detect his or her face. Although the above operations show a bright future for the application of facial recognition technology, in effect, it is not a perfect technology. Stated thus, the facial recognition technology does not achieve perfect accuracy, when it is applied in law enforcement, especially when it identifies suspects, it cannot provide 100% correct results. Hence, facial recognition can not be widely used in law enforcement.

Topic sentence: Secondly, discrimination is also a weak point that facial recognition technology should not be used in law enforcement because it does not accurately detect people of color. As everyone knows, facial recognition technology is developing rapidly recently. Evidence & citing: In an experiment that tested this technology to identify different ethnic groups, the results showed that the accuracy of this technology was so high in identifying white people, even almost flawless. When it comes to identifying people of color, however, the accuracy has obviously decreased. The darker the skin color of the tested person, the lower the accuracy of the face recognition technology. Even, the error rate can reach 35%. Moreover, the experiment of Joy Boulamwin had shown that when facial recognition systems work, testing people of color is more likely to get wrong results than testing white people. When Joy Boulamwini studied at MIT in 2019, she discovered the facial recognition technology was really hard to test her face unless she wore a white mask, simply because she had dark skin. In many cases, facial recognition technology is not objective. According to Because We Can Free Our Machines of Bias, Joy Boulamwini showed that “facial recognition systems sold by the largest technology companies, like IBM, Microsoft, and Amazon, have discrimination. Even, very well-known faces -- the faces of Oprah Winfrey, Michelle Obama and Serena Williams can not be recognized successfully by facial recognition technology from the most recognizable companies”. Commentary: Therefore, when realizing that the facial recognition technology has the problem of discrimination, its application in the police department needs to take more consideration. For this reason, the facial recognition technology with prejudice is hard to get the correct decision when it recognizes the criminals of colored people.

Finally, the third drawback of facial recognition technology is the menace to an individual's privacy. If the government wants to fully apply facial recognition technology, it may require enterprises to install an electronic surveillance countrywide to scan the faces of customers and employees. Some people mentioned the possibility of personal information being abused, since cameras and monitors can easily be connected to the Internet, tracking individuals everywhere, and facilitating the government to grasp the situation and control social order. Citizens are also concerned that businesses will begin to create data documents about their clients and workers, and then share personal data with other businesses, but most importantly, personal information should be classified. It is to be mentioned that with the development of technology and the ability to accumulate an overwhelming amount of data, police department encounters the risk of damage to storing information. Commentary: Every piece of information collected, if leaked, may be used maliciously. Especially, the information collected by facial recognition technology during enforcing the law has a high risk of hacking and data abuse. For example, it may be used by terrorists or other interested organizations. Since the data collected by the government is crucial, the government database is often the biggest target of hackers. So, the threat of information theft to citizens is huge and unimaginable. For instance, in 2014, the United States Office of Personnel Management, in which the data of about 22 million people was leaked, including work history, family members' names, fingerprints. Such as the OPM hacking incident, which is an example of when the public trust in the government was damaged, and the police department and other departments were violated by security loopholes. The consequences are irreparable and unimaginable. People are just afraid that facial recognition technology collected information can be used to violate privacy.

In addition, a Georgetown Law Study indicated that a majority of American adults are included on the police facial recognition database. Since citizens think that the monitors equipped with facial recognition system can capture their all actions in daily life, they are not only worried that they are being monitored all the time but also concerned that the consequences of the application of facial recognition technology in the police department without rules. In fact, there are little or no laws that can regulate the application of facial recognition technology. This means that facial recognition technology can easily intrude on peoples’ privacy.

Conclusion

Conclusion paragraph: In conclusion, this paper discusses the disadvantages of the application of facial recognition technology in law enforcement from three aspects: poor reliability, race discrimination, and privacy-intrusive.

Facial recognition is not always able to accurately identify a criminal, it is affected by the camera angle, human appearance, and even a database. In particular, if the information of the investigated criminals is not in the database and the facial recognition technology can not compare, the final result is that the facial recognition technology provides a similar face, that will lead to the error of arresting criminals and greatly reduce the rate of solving cases. Moreover, facial recognition technology has a bias problem, it is not good at recognizing faces when identifying people of color. Facial recognition technology is only used to recognize white people with high accuracy. When identifying people of color, it has a high error rate. So if the police need the help of a facial recognition technology to identify a criminal of color, the result would be unsatisfactory. Last but not least, facial recognition technology is considered a violation of privacy because citizens think that they are exposed all the time and they do not want to be watched for everything they do. They not only think they have no privacy but also are worried that their personal information will be exposed, since there is barely any law to regulate the application of facial recognition technology in law enforcement.

I suggest the application of facial recognition technology in law enforcement needs the combination of the technology and people to achieve the highest efficiency, and set stricter laws or policies to protect citizens' privacy.

Reference

1. BBC News. (2018). Chinese Man Caught by Facial Recognition at Pop Concert.
2. Bhatia, R. (2013). “Biometrics and Face Recognition Techniques.” International Journal of Advanced Research in Computer Science and Software Engineering 3(5), 93-99.
3. Boulamwini, J. (2019). Because We Can Free Our Machines of Bias. TIME Magazine. 193(6/7), 67-68.
4. Carter, M.A. (2018). Facing Reality: The Benefits and Challenges of Facial Recognition for the NYPD.
5. Eggers,W D. (2016). Delivering on Digital: The Innovations and Technologies That Are Transforming Government.
6. Jennifer, L. (2018). Face Off: Law Enforcement Use of Face Recognition Technology.
7. Lohr, B. (2018). Facial Recognition Is Accurate, if You’re a White Guy.
8. Mallon, B. (2003). “Every Breath You Take, Every Move You Make, I'll Be Watching You: The Use of Face Recognition Technology.” Villanova Law Review, 48(3). 955-988.
9. McAllister, D. (2007). “Law Enforcement Turns to Face-Recognition Technology.” Canada: Information Today, 50-51.
10. N.A., (2001). What's to stop commercial companies from acquiring these public records and selling them.
11. N.A. (2017). Face Recognition Policy Development Template For Use In Criminal Intelligence and Investigative Activities.
12. Phillips, P. J.& Beveridge, J. R. (2011). An Introduction to the Good, the Bad, & the Ugly Face Recognition: Challenge Problem. Santa Barbara, CA: IEEE.
13. Ringrose, K. (2019). “Law Enforcement's Pairing of Facial Recognition Technology with Body-Worn Cameras Escalates Privacy Concerns.” Virginia Law Review Online, 57-66.
14. Wendt, R. (2019). Face the Future of Surveillance: A ban on face recognition hampers law enforcement for now, while paving the way toward its use in the future. California: Cygnus: Business Media.
15. Woodward, J, D. (2003). Biometrics: A Look at Facial Recognition. Santa Monica, CA: RAND Corporation.