Facial Recognition Technology - Ethical Perspectives

of one

Facial Recognition Technology - Ethical Perspectives

While many are excited about the potential applications of facial recognition technology, there are also a number of ethical concerns relation to it, such as developmental bias, informed consent regulations, data storage, incidences of mistaken identity, and government surveillance, among others.

1) Developmental Bias

  • One of the primary ethical concerns relating to facial recognition technology is that of developmental bias.
  • When developing facial recognition software and technology, creators focus on (among other things) "mapping ability and positive identification tools by practicing on massive data sets. Unfortunately, many of these data sets are not diverse... [and] overrepresent Caucasian individuals as well as men, people without disabilities and middle-aged adults".
  • This can mean that, in practice, facial recognition software may tend toward racist, sexist, and/or ableist stereotypes simply because the software has been largely trained on white, male, and able-bodied faces. Ultimately, software which has been trained on a non-diverse set of images "won’t give you as accurate results as when training on a more diverse dataset that better represents real-world conditions".
  • For example, software trained only on white faces may misidentify an innocent person of color as a suspected criminal, simply because it doesn't recognize individual faces of color as well as white ones. In fact, research from an FBI co-authored study has found that "police facial recognition systems don’t work as well on black people’s faces". This can contribute to a justice system which already unfairly over-polices black people.

2) Informed Consent

  • Relating to the use of facial recognition technology in healthcare specifically, another ethical concern is that of informed consent.
  • "As FRT is increasingly utilized in health care settings, informed consent will need to be obtained not only for collecting and storing patients’ images but also for the specific purposes for which those images might be analyzed by FRT systems. In particular, patients might not be aware that their images could be used to generate additionally clinically relevant information."
  • One major concern is that, "while FRT systems in health care can de-identify data, some experts are skeptical that such data can be truly anonymized". This means that while a patient may be told their data is private or anonymous, in practice it might not actually be.
  • For example, some medically-used facial recognition datasets need to be continually updated in order to ensure accuracy. This might mean including images of patients who have already been diagnosed, "for which informed consent is not regarded as necessary". Given HIPPA laws concerning privacy around medical data and diagnosis information, this poses an ethical concern relating to facial recognition technology.

3) Data Storage

  • Another ethical concern relating to the use of facial recognition technology is that of whether or not it should be legal to store the biometric data of another person.
  • "This can be a problematic area of consent, because a person’s biometric data could be used in innumerable nefarious ways, or ways that infringe upon their privacy, such as accessing their devices or tapping into their personal data, from where they went to high school to what their favorite type of ice cream is (assuming you have the right tools)."
  • While there are some states in the United States which have begun to draft legislation concerning this topic, many other states and countries have not. Additionally, those laws which have been drafted tend to apply to businesses, rather than to individuals.

4) Mistaken Identity

  • One of the most obvious applications of facial recognition technology is in law enforcement, to help the police or other judicial systems find and detail criminals, even if they've tried to disguise themselves.
  • However, while this is an exciting development at face value, the potential for mistaken identity is significant and several instances of this happening have been recorded.
  • This is a potential result of developmental bias, in that a technology which has been trained only on one specific kind of face will have trouble recognizing differences in other types of faces.

5) Surveillance without Justification

  • Another significant and common concern relating to the use of facial recognition technology by law enforcement is that of undue surveillance of innocent citizens.
  • For example, police could survey and track faces from a peaceful protest. That data could later be used to track and monitor the protests, regardless of the fact that it was peaceful and no crime was committed.
  • "With few scientific standards or government regulations, there is little preventing police departments from using facial recognition to target immigrants or identify participants in a political protest, critics say."
  • In fact, the ACLU has indicated that this concern — that of governments using FRT for "general, suspicionless surveillance systems" — is the single greatest ethical concern relating to the use of facial recognition technology.

6) Customer Service

  • Moving away from law enforcement, another ethical concern relating to facial recognition technology is that of its use in customer service settings, such as on in-store customers.
  • Should businesses be able to use facial recognition technology, they could implement it to target and identify certain customers.
  • For example, they would be able to keep "records on individual consumer accounts and tracking details like what size clothing they wear, what their product preferences are, where they live, and what demographic categories they fall into".
  • While this could result in better customer service since employees will already know what you're looking for when you enter a store, it could also result in worse service for customers that the technology determines won't spend much money.

7) Account Hacking

  • One way facial recognition technology has already been implemented has been on smartphone locks.
  • Specifically, iPhone and iPad allows users to set up a lock screen which opens based on facial recognition rather than a code or touch pattern.
  • Given that "there are already many ways to fool facial recognition technologies, with photos, prosthetics, and special “noise” layers that function kind of like Instagram filters", another ethical concern relating to the use of FRT is that of hackers using photos from social media or other sources to "trick" technology into letting them into a locked source.

8) School Monitoring

  • In China specifically, facial recognition technology is being applied in schools, where it's used to track student facial expressions during class, in order to determine whether or not they're paying attention. School officials then combine that data with student test scores to determine their overall performance in individual classes.
  • As some have pointed out, while the ability to track student performance may sound good at face value, the use of FRT in this way could pose "a threat to the ability of expressing emotion through facial expression freely and the ultimate right of a human being to have private thoughts".
  • The same software has also been used in Chinese schools to track attendance and other metrics in the classroom. However, the country's Ministry of Education announced in September that it intends to pass guidelines and regulations aimed at curbing how FRT is used in the school environment based on ethical concerns which have been raised.

9) ACLU Ethical Framework

  • The ACLU has created and submitted a potential ethical framework to be applied to the use of facial recognition technology.
  • Among other things, the framework requires that FRT users receive informed and voluntary consent from individuals before using their face in a database and keep those users and their consent forms updated concerning any change to use.
  • The framework also includes considerations focused on sharing, access, misuse, security, accountability, government access, alternatives, children and teens, and transparency.

10) European Union Ethical Regulations

  • Similarly, the EU has begun drafting a new regulation concerning the use and application of facial recognition technology.
  • The new regulations will ensure that "citizens of the European Union will have clear rights over how facial recognition technology companies use their data, as officials plan to curb creeping surveillance of ordinary people".
  • "The non-consensual harvesting of biometric data is already outlawed under EU data protection regulations"; however, these new regulations would apply more specifically to the use of artificial intelligence (AI).