When the COVID-19 pandemic shuttered physical classrooms, students arrived home to numerous Zoom or Google Meet invitations to attend their newly established virtual classes. Now that remote education has become ubiquitous (despite optimistic plans for in-person classes in Fall 2021), many educational institutions have turned to the trendiest technological advances to monitor classroom activity, particularly when it comes to exam administration. In May 2020, facial recognition and similar artificial intelligence (AI) systems were more prevalent in classrooms in China than in the U.S. However, U.S. classrooms have now quickly pivoted to these technologies. The transition has been quite abrupt in some cases, so much so that potential invasions of privacy have been overlooked. Although outrage at the K-12 level has yet to morph into concrete action, some college students have been protesting their universities’ use of the allegedly invasive technology. The issue has also been a significant concern surrounding administration of the bar exam. In one case, it has led to litigation.
While facial recognition and AI software may be used for building security and classroom monitoring when students are in physical classrooms, their primary purpose in the recent shift to remote education has been to detect cheating on exams. Many of these programs use facial recognition technology to ensure that the enrolled individual is sitting for a given exam or, alternatively, to determine whether that person is looking off-screen at impermissible open-book material for extended periods of time. They also often record keystroke patterns to determine if a student is typing their answers or simply copying and pasting from a separate document. A more controversial issue is that remote proctoring software may also be used to access a student’s camera and/or microphone, which effectively gives a school access to their home. This access raises the concern that companies are collecting personally identifiable information (PII), including biometric data (such as eye movement and facial expressions) in conjunction with names, student ID numbers, IP addresses and other similar information. In addition to these companies aggregating this PII, privacy experts are concerned with breaches of their databases. Unfortunately, this is not a theoretical concern. In August 2020, hackers leaked the information of over 400,000 registered users of online proctoring company ProctorU.
There has been further concern with respect to discipline as a result of errors from remote proctoring software. At Dartmouth’s Geisel School of Medicine, several students were recently accused of cheating on remote exams after school officials reviewed their activity on the remote platform Canvas, alleging that they viewed prohibited material during the administration of their exams. According to one of these students who alleges the determination was a result of a system error, the school has suggested that students admit to their misconduct despite pleas of innocence. Any such admission of a transgression could have a negative impact on a medical student’s academic record and future job prospects. Of course, these issues are not simply limited to the medical field — virtually anyone seeking a job may also be vulnerable. This unnecessary stress at Dartmouth has led students to protest on campus and faculty members to write letters of concern to the administration, attracting national attention.
Some of these cheating accusations may indeed be valid, but the reality is that nothing is certain. There are students who may live in close quarters in low-income or even just group housing, where it is difficult to avoid other family members or housemates entering the room in which they are taking an exam. Certain AI programs may determine that these individuals are providing prohibited assistance. Similarly, students with an unreliable Internet connection may trigger proctoring software algorithms that determine that they have been leaving the room or accessing impermissible information. Per the Dartmouth incidents, this has caused trouble even for those with completely stable Internet connections. It is entirely possible that students simply have difficulty focusing on their screen for an extended period of time. Maintaining eye contact in a “Zoom class” is enough of a problem, especially for students with diagnosed attention span issues. Coupled with testing jitters, facial recognition detection of “cheating” when looking off-screen only adds to the anxiety.
Furthermore, criticism of these platforms has been met with fervent suppression by the providers. After college student Erik Johnson posted portions of online proctoring company Proctorio’s software on Twitter, the company used the Digital Millennium Copyright Act (DMCA) notice-and-takedown policy to remove the tweets that allegedly violated its exclusive right to display its copyrighted work. In response to Proctorio, the Electronic Frontier Foundation (EFF) filed a complaint on behalf of Johnson, seeking a declaratory judgment, injunctive relief and damages in an effort to “finally quash a campaign of harassment designed to undermine important concerns about software used by universities around the United States to monitor student activity.” Johnson claims that not only did he post portions of the code to expose privacy and security issues, but he also chose excerpts that were publicly available, which would not constitute copyright infringement.
Johnson argues that his posts amount to fair use of the copyrighted software code because he is merely using it as evidence to support his criticism of Proctorio’s potential breaches of privacy. In his Twitter thread, Johnson also mentioned an instance in which the CEO of Proctorio released a student’s chat log involving criticism of the company on Reddit. Furthermore, Johnson highlighted the inequity by stating that “[f]orcing a student to have a private, quiet room with a quality webcam, microphone, stable internet connection, and computer can be nearly impossible for some.” While this may appear to be vigorous litigation protecting student privacy and security, this complaint did not set forth any legal claims outside of copyright law. Therefore, a victory for Johnson could ultimately be confined to permitting further criticism of the company’s practices.
Moreover, this is not the first time that Proctorio has been an ardent defender against its critics. In October 2020, the company sued a learning technology specialist in Canada for copyright infringement, alleging unlawful appropriation of their confidential training videos in several tweets that criticized the company’s practices. The CEO of Proctorio has publicly stated that the company was merely defending its intellectual property. On the other hand, he has engaged in various Twitter arguments and requested retractions of articles that place a spotlight on the company’s questionable conduct. The aggressive response to any opprobria has left many in the tech and legal community to wonder how these copyright infringement lawsuits may proceed, and if other online proctoring companies will follow Proctorio’s lead.
Congress has also played a role in this arena. In December 2020, U.S. Senators Ron Wyden (D-OR), Chris Van Hollen (D-MD), Tina Smith (D-MN), Elizabeth Warren (D-MA) (a former Harvard Law School Professor) and Cory Booker (D-NJ) authored a letter to the three most prominent online testing companies — Proctorio, ProctorU and ExamSoft—in an effort to address these issues directly. The letter noted the problems exacerbated by economic inequity, racial and gender bias with respect to facial recognition technology, discriminatory effects against individuals with disabilities and the broader privacy and safety concerns posed by constant monitoring. This panel requested responses from the three companies, emphasizing the potential violations of the Family Educational Rights and Privacy Act (FERPA), the Higher Education Act and Title II of the Americans with Disabilities Act. Although the cases currently involving Proctorio are focused on copyright infringement, it remains to be seen if these federal laws are invoked in potential future litigation.
Lutzker & Lutzker will continue to monitor the cases involving Proctorio, provide updates as they progress and inform you of further discourse about online proctoring software using AI and facial recognition technology. Please contact us if we can be of help in analyzing and addressing copyright and privacy issues related to the use of such software.