A Closer Look: Assessing the Privacy Costs of School Safety Proposals
In an effort to prevent school violence, a number of communities and educators are exploring new policies and technologies that are intended to help identify and flag potential threats. However, each of these efforts has serious and lasting implications on student privacy and data security. To help raise awareness of the challenges and opportunities presented by these school safety policies, the Future of Privacy Forum (FPF) has developed a weekly brief that will offer a closer look at a few of the most prevalent technologies along with useful resources and articles to further inform readers. Don’t miss our previous issue briefs on school network monitoring and social media monitoring.
Issue Brief: Facial Recognition Technology
Under immense pressure to address school shootings and other safety threats, a few schools are adopting facial recognition systems. Most facial recognition technology utilizes software and cameras to conduct live biometric scans of students, teachers, parents, and any other school visitors, and then cross-check that information against a database of individuals not welcome on school property. Although the desire to provide the highest levels of protection for students and school personnel is well-intentioned, there is no evidence that the use of facial recognition systems will actually make schools safer.
While a few advocates worry that banning facial recognition in schools could limit “creative and positive” uses that could potentially keep students safe, most groups agree that facial recognition is “uniquely ill-suited to respond directly to school shootings.” Analysis shows that the vast majority of school shooters are current or former students, and therefore unlikely to be flagged in a database. And tragically, even if a facial recognition system does correctly identify an attacker, the timeframe of most school shootings is so narrow that an alert would not allow first responders enough time to intervene. Furthermore, purchasing and maintaining these systems is often cost-prohibitive, with typical expenses for a school of about $100,000 annually.
Accuracy and abuse are additional major challenges: tests have shown that some systems have trouble matching accuracy rates for identifying women, people of color, and young people. Schools may be tempted to use the technology for discipline, another issue that exacerbates racial inequities.
Furthermore, the loss of privacy and constant surveillance makes students feel like “perpetual suspects” and erodes trust between schools and students and their families, especially when use expands beyond safety to a “convenient way to monitor students.” This effect can be particularly chilling for undocumented students and families. Schools that implement facial recognition may also face backlash from parents, volunteers and staff who don’t want to be involved in such a system, decreasing the willingness of people to work or volunteer at the school.
However, these concerns do not mean that all biometric technologies, including those that may use facial recognition, are problematic: schools may implement facial categorization technologies that, if banned outright, could prevent or compromise current services to students. For example, schools may currently use biometric software that does not identify individuals but measures facial expressions, voice data, or gait analysis in order to help students in special education, or biometric systems based on fingerprints and palm prints for lunch-line efficiencies or attendance reporting.
School districts and communities considering the broad use of facial recognition technology for threat detection must weigh not only the significant financial costs and questionable accuracy and effectiveness of the technology itself, but also the serious consequences on student privacy, equity, and broader school and community culture that would likely result.
What They’re Saying: News and Research
- Wired: The Delicate Ethics of Using Facial Recognition in Schools
- Wired: In the Face of Danger, We’re Turning to Surveillance
- New York Times: Spying on Children Won’t Keep Them Safe (Opinion)
- District Administration: Biometric security boosts school safety and efficiency
- Wall Street Journal: One School District’s Security Upgrade: Facial Recognition, Tracking IDs and AR-15s
- Fed Scoop: Facial recognition algorithms are getting a lot better, NIST study finds
- New York Times: She Was Arrested at 14. Then Her Photo Went to a Facial Recognition Database
- Bloomberg: Researchers Call for Regulation of Facial Recognition, AI Tools
- Human Rights Watch: Facial Recognition Technology in US Schools Threatens Rights
- ACLU of New York: Facial Recognition Cameras Do Not Belong In Schools
- Federal Trade Commission: Facing Facts: Best Practices for Common Uses of Facial Recognition Technologies (2012)
- National Telecommunications and Information Administration: Facial Recognition Technology Best Practices
- Facial Recognition Video Series
Hear from FPF’s expert on biometrics, Brenda Leong, in a series of short video clips on the accuracy of facial recognition technology, the associated privacy risks, and if it should be used in schools.
- Facial Recognition Systems in Schools
FPF held a 1-hour webinar for schools with experts from FPF and the Rochester Institute of Technology on the rationale, risks, and recommendations regarding the use of facial recognition in schools
- Letter to New York State Legislature on the Use of Facial Recognition Schools
In June, FPF wrote in support of a well-crafted moratorium on facial recognition systems for security uses in public schools, while cautioning against overly broad bans or language that might have unintended consequences on other security programs.
- Principles for Student Safety, Privacy, & Equity
Written by 40 education, privacy, disability rights, and civil rights organizations, these principles are designed to protect students’ privacy, dignity, and right to an equal education.
- Privacy Principles for Facial Recognition Technology in Commercial Applications
Published in September 2018, these principles define a benchmark of privacy requirements for commercial situations where technology collects, creates, and maintains a facial template that can be used to identify a specific person. The publication also includes Potential Privacy Challenges Associated With Facial Recognition (page 14).
- Understanding Facial Detection, Characterization, and Recognition Technologies
This infographic breaks down key characteristics of the the different types of facial detection, characterization and recognition technologies.