.jpg)
Proctoring software is common in remote exams, but it comes with a long list of student privacy concerns. Some institutions use it only where there is no workable alternative. Others keep it switched on for digital exams even when students are sitting in a supervised room.
The reason this is so sensitive is simple: online proctoring can ask students to share far more than just their exam answers. Depending on the setup, it may capture their screen, webcam, microphone, room, ID document, and behavior during the exam. That is a lot of personal data to collect in a high-pressure setting where students often have little real choice but to accept the conditions.
Institutions need to question whether the use of proctoring software is necessary and proportionate in the specific exam setting. Can they show a valid legal basis, a genuine need, and proof that there is no less intrusive way to protect the exam?
There are situations where online proctoring can be defensible, especially for genuinely remote exams with no feasible form of human invigilation. There are also situations where it becomes much harder to justify, especially when students are already sitting in an exam hall with supervisors present.
This article covers:
- the difference between proctoring and lockdown solutions
- what courts and regulators in the Netherlands, Portugal and the United States say about these privacy concerns
- when online proctoring is justified, and the conditions that come with it
- when it becomes an invasion of privacy, and what to use instead
The difference between proctoring and lockdown software
People often use the word ‘proctoring’ as a catch-all for any exam security tool. That blurs an important line, and the rest of this article depends on it.
Proctoring watches. It uses webcam, microphone, and screen recordings to monitor a student during the exam, and stores that footage for a human or an algorithm to review.
Lockdown software prevents. It restricts what a device can do during an in-person exam, closing off unnecessary applications, browser tabs, and AI tools, without recording anyone in the process.
What courts and regulators say about online proctoring privacy concerns
The Netherlands: is there a less intrusive alternative?
Two student councils took a university to court over its use of a proctoring tool, arguing it was too invasive to be lawful. They lost, twice: before the District Court in 2020, and on appeal in 2021. The Court of Appeal accepted it because the lockdown during the pandemic had closed the campus and no adequate alternative was available. It described the software as standing in for the physical invigilator.
In other words, the court allowed the software specifically because an in-person invigilator was not an option at the time, not as a general green light for surveillance during any digital exam.
Portugal: automated fraud scores
In 2021, Portugal's data protection authority, the CNPD, reviewed a university's use of a proctoring tool for remote exams. The tool tracked a student's mouse and keyboard use together with their facial and body movement, and turned all of that into a score for how likely they were to be cheating. The CNPD ordered the university to stop using it and delete everything already collected, finding the setup broke basic data protection rules.
A teacher still made the final call on whether to fail a student for cheating, so on paper a human was in charge. But the CNPD said that didn't count as a real human decision, because nobody gave teachers clear guidance on how to read the score, so a teacher could just wave through whatever the tool produced.
The United States: room scans
In the US, one federal court drew a hard line around room scans. In Ogletree v. Cleveland State University, a federal judge ruled that scanning a student's bedroom before a remote exam was an unreasonable search of his home under the Fourth Amendment.
When online proctoring can be justified
If students cannot reach campus and no feasible form of human invigilation is available, online proctoring may be proportionate. But even then, the institution still has to show why less intrusive alternatives would not protect the exam sufficiently.
When proctoring, take the following into account:
- base the usage on a proper legal basis, not student consent, since consent given under pressure to pass a course is rarely freely given
- collect the minimum amount of data, and drop room scans and biometric checks unless you can prove they are necessary
- set short data retention periods and delete recordings promptly
- keep the data within a jurisdiction that protects it properly, or put genuine safeguards in place if it has to cross borders
- be transparent with students about what is being recorded
- if the tool produces an automated fraud or suspicion score, make sure a human can genuinely review and override it, not just confirm it
Meet the above criteria and proctoring becomes much easier to defend. Miss them, and the privacy risk rises quickly.
Why proctoring can’t be justified in an in-person setting
In an in-person exam you already hold the less intrusive option: an invigilator. That makes recording every student's face, screen and behavior hard to call necessary, which is the exact necessity the Dutch ruling rested on. Read in reverse, that reasoning says the justification for using proctoring tools falls away as soon as a supervisor is present.
A quick test you can apply to find out if your institution should use a proctoring tool
You can settle most cases with three questions.
- Can a human invigilate this exam in a room or hall? If yes, proctoring is hard to justify, and you should use cheating prevention software instead.
- Is the exam genuinely remote with no feasible way to supervise? If yes, proctoring may be proportionate.
- Are you reaching for room scans or biometric verification? Treat those as a red flag almost anywhere, and drop them unless you can prove they are necessary.
- Does the tool generate an automated suspicion or fraud score? If yes, make sure a human meaningfully reviews it before any consequence follows.
If your exams are sat on campus, you are almost always in the first case. That is where most institutions are, and it points away from surveillance entirely.
The proportionate route for in-person exams
So what do you use when proctoring software is the wrong solution? For an in-person digital exam, the challenge of cheating prevention was never supervision. You have a person in the room for that. The problem is all the cheating routes on the device. When a student takes an exam digitally, they can open AI tools like ChatGPT or Claude, switch to another application, or look something up, and one invigilator cannot watch that across thirty screens at once.
The proportionate fix is to control what resources the student can access on their device during the exam, without recording anyone. This is the difference between monitoring and prevention. Proctoring monitors and records so you can judge behavior later. Prevention avoids the most intrusive data: webcam footage, microphone recordings, room scans and behavioral surveillance. It may still create technical and security logs, but those are different from recording the student throughout the exam.
Why WHU swapped proctoring for the Safe Exam Workspace due to privacy concerns
WHU – Otto Beisheim School of Management ran into this exact problem. When campuses reopened after the pandemic, WHU kept its remote proctoring tool running for on-campus digital exams. It didn't go well.
The tool flagged behavior with AI, and then someone had to review the footage after the exam to work out whether it actually meant anything. "The workload came after the exam, and even then, it was often unclear whether someone was actually cheating," says Thorsten Leich, Associate Director IT Solutions.
There were also privacy concerns. "WHU didn't want to surveil students like suspects," says Vincent Meertens, Program Director BSc/MSc. "It's about ensuring that students can't access outside sources, without overdoing it."
WHU switched to Schoolyear's Safe Exam Workspace. Students bring their own laptops, and the device is locked to what the exam allows: full-screen mode, other applications closed, only pre-approved websites and files reachable. Instead of recording potential cheating, cheating is prevented.
