A recent article in The New Yorker asks, “Is Online Test-Monitoring Here to Stay?”.

Unfortunately, the answer is probably yes, especially since the broader student surveillance industry is rapidly expanding.

Because large numbers of college teachers don’t trust their students (“In a survey… conducted early in the pandemic, ninety-three per cent expressed concern that students would be more likely to cheat on online exams.”), colleges at the start of the shutdowns quickly signed up for systems that would automatically proctor online testing for them.

Software that uses the student’s own camera, microphone, and computer to spy on them.

Proctorio, which operates as a browser plug-in, can detect whether your gaze is pointed at the camera; it tracks how often you look away from the screen, how much you type, and how often you move the mouse. It compares your rate of activity to a class average that the software calculates as the exam unfolds, flagging you if you deviate too much from the norm. Meanwhile, Proctorio is also monitoring the room around you for unauthorized faces or forbidden materials. At the end of the exam, the professor receives a report on each student’s over-all “suspicion score,” along with a list of moments, marked for an instructor to review, when the software judged that cheating might have occurred.

That “suspicion score” is generated by a proprietary algorithm, which reflects the biases of both the programmers and the data they used.

Other anecdotes call attention to the biases that are built into proctoring programs. Students with dark skin described the software’s failure to discern their faces. Low-income students have been flagged for unsteady Wi-Fi, or for taking tests in rooms shared with family members. Transgender students have been outed by Proctorio’s “ID Verification” procedure, which requires that they pose for a photograph with an I.D. that may bear a previous name. In video calls with live proctors from ProctorU, test-takers have been forced to remove bonnets and other non-religious hair coverings—a policy that has prompted online pushback from Black women in particular—and students accessing Wi-Fi in public libraries have been ordered to take off protective masks.

The article spends a lot of time with the CEO of Proctorio who has gained an infamous reputation as first-class jerk, to put it mildly.

But the most salient difference between Proctorio and its peers may be Olsen. Last April, he pressured Hybrid Pedagogy to retract an article by Shea Swauger, a librarian and researcher at the University of Colorado Denver, that was critical of his company. (The journal refused.) In September, Olsen filed a lawsuit against Ian Linkletter, a learning-technology specialist at U.B.C., claiming copyright infringement after Linkletter tweeted publicly accessible company videos. (Linkletter told me that he has exhausted his life savings fighting the suit.) In October, Olsen pursued takedowns of a computer-engineering student’s tweets and Pastebin posts that pointed out potential security weaknesses in Proctorio’s handling of students’ private data. (The student is now suing Proctorio, alleging that the company misrepresented its copyright claims.)

Despite the biases hard-wired into these systems, and the lack of evidence for both the need and effectiveness, their use will expand. Someone – or something – must be vigilant watching for cheaters. Professors can’t possibly be expected to do the hard work necessary to create assessments that better reflect student learning.

However, even if you’re not involved in college teaching, remember: as online K12 instruction grows, even beyond the pandemic, public, private, and charter schools are probably going to join this student surveillance party.

Think of that picture (from an article on standardized testing in The Atlantic) as what Proctorio and their competitors are trying to replicate online.