A webcam policy for remote exams settles whether a camera in a student's home is taken as a fair condition or as an intrusion. The studies below suggest the objection is seldom to the lens alone.
Surveys of proctored exams are in webcam proctoring; the notice for a quiz with no camera is in what to tell students before a monitored online test.
EdProcto publishes this guide, and its own integrity settings act in the browser. Every source was read on 10 October 2026; "abstract read" marks a study read only in abstract. Nothing here is legal advice, and no wording is claimed to comply with any law.
Why do students object to a webcam in a remote exam?
For four recurring reasons: the home on view, not knowing who watches and keeps the recording, bandwidth and devices, and anxiety. Most figures below come from cameras in live online classes, where a student could switch off and say why. An exam removes that choice.
The home on view
Castelli and Sarvary (2021) surveyed 276 undergraduates in a biology laboratory course at Cornell University in spring 2020. Asked why they had left the camera off, 41% ticked concern about their appearance, 26% other people being seen behind them and 17% their physical location being seen. Among students from underrepresented minorities the last two figures were 38% and 26%, against 24% and 13% for the rest: exploratory comparisons, the authors say.
Rajab and Soheib (2021) surveyed 319 medical students at a Riyadh university, where 80.6% lived with family; about a quarter of those invited answered. Cameras off was the preference of 292 (91.5%), and of those 258 (88.4%) gave privacy as a reason. Comments named family members in the background.
Who watches, and who keeps the recording
Alessio and Messinger (2021) drew replies from 150 faculty and 78 students at a large public university in the American Midwest. Online proctoring was an invasion of privacy to 70% of the students and 53% of the faculty. Given the choice, 2% of students and 46% of faculty were at all likely to pick proctoring software. Only 8% of invited students replied.
On who opens and keeps a recording, the clearest account we read is the Ontario report below.
Bandwidth and devices
At Cornell, 22% said their internet connection was weak and 2% that the webcam was not working. In Riyadh, 94 of the 292 (32.2%) named technical issues and 24 (8.2%) a lack of technology. In a class a frozen picture costs nothing; in an exam it can look like an absence.
In an online graduate computing programme surveyed by Duncan and Joyner (2022; abstract read), students and teaching assistants were generally tolerant of digital proctoring, with concerns about privacy, equity and technical difficulties.
Anxiety
Anxiety was the second most common reason in Rajab and Soheib (2021): 188 of the 292 (64.4%). At Cornell, 17% felt that everyone was looking at them the whole time. Exam scores under a live camera are in how much proctoring a quiz needs.
What makes monitoring feel fair to the people monitored?
Four things, in the studies we read: notice in advance, a stated reason, a say or an alternative, and limits on what is watched. Two are workplace studies, so the step to an exam is our inference.
- Notice, and a reason. Hovorka-Mead et al. (2002; abstract read) studied seasonal employees, school and college students, monitored by video camera. Those told in advance judged the monitoring fairer. In a scenario experiment, any justification, strong or weak, produced more belief in the procedure's fairness than none.
- A say, and limits. Alge (2001; abstract read) had temporary workers do computer tasks under varying surveillance. Monitoring only the activity relevant to the job, and giving those monitored some input, reduced the sense of invaded privacy and raised perceived fairness.
- A purpose students share. Whitelock et al. (2020; abstract read) report trials of identity tools, face and voice recognition among them, with 4,058 students at seven institutions. Over half responded positively, and over 70% agreed that the tools were "to ensure that my examination results are trusted".
Castelli and Sarvary (2021) advise offering alternatives to a required camera, and Alessio and Messinger (2021) that such policies be written with student input. We found no study that tested one exam policy against another, and none that varied how long a recording was kept.
What have regulators and courts said about remote proctoring and room scans?
Case by case, they ask for much the same list: need, notice, an alternative and limits on recording. Each page is reported as its own, under its own law; which rules bind you is for your lawyer or data officer.
| Body and country | Date | What its own page says | Checked on |
|---|---|---|---|
| Information and Privacy Commissioner of Ontario (2024), Canada | 28 February 2024 | On one university's use of Respondus software: proctoring certain exams online was an authorised activity. The notice fell short, since students had to piece it together from several pages, and the vendor's use of recordings to improve its product was not something students would expect. Recommended: one plain notice, yearly deletion at the least, an opt-out for in-person invigilation, an informal way to challenge a flag | 10 October 2026, archived copy |
| Landesbeauftragter für den Datenschutz und die Informationsfreiheit Baden-Württemberg (2021), Germany | 17 July 2021 | Under that state's higher education law: consent is as a rule no basis, since it is not freely given to a public body; students are told everything in advance; taking part is voluntary, with an in-person exam on the same date; nothing is recorded; no tracking of eyes, head or body | 10 October 2026 |
| Oberverwaltungsgericht für das Land Nordrhein-Westfalen (2021), Germany | 4 March 2021 | Refused a student's urgent application against a distance university's rules, under which video, sound and screen were recorded in home exams and deleted afterwards unless the invigilator had noted an irregularity or the student asked for a review. Lawfulness could not be settled in urgent proceedings | 10 October 2026 |
| Agencia Española de Protección de Datos (2025), Spain | 3 June 2025 | Its summary of a sanction over a university's remote exams: facial recognition and a second camera were imposed with no valid alternative. Consent was invalid because refusing meant losing the assessment. The resolution can be appealed | 10 October 2026 |
| Persónuvernd (2020), Iceland | 17 November 2020 | Guidance on distance teaching: where exams are sat with cameras on, students are unlikely to be able to consent freely. They must be told in simple language, and the school goes no further than needed | 10 October 2026 |
What the pages say about room scans
Little: two of the five touch them.
- Germany. Baden-Württemberg's guidance says a camera pan through the student's room is not permitted under that state's law, even where cheating is suspected.
- Spain. The sanctioned system watched the student's surroundings through the second camera, with software looking for other people and forbidden objects.
An Ohio federal court's ruling on a room scan, and decisions from Italy, Denmark and the Netherlands, are in the webcam guide linked above.
Where they differ
On recording itself: Baden-Württemberg's law forbids it, the rules the North Rhine-Westphalia court let stand deleted each recording after the exam unless something was noted, and Ontario asked for yearly deletion at the least. No retention period carries from one place to another.
What should a webcam policy for remote exams say?
Six things: what is recorded, why, who sees it, how long it is kept, what the alternative is, and how to object. One decision comes first: the table's first row.
| The policy states | Why it belongs | Say this, not that |
|---|---|---|
| Whether a camera is used at all | The lightest record that does the job | "This exam logs browser events only", where true: EdProcto logs them and takes no picture |
| What is recorded | The home is the first objection | "Picture and sound, never the room", not "your session may be monitored" |
| Why this exam needs it | Any reason beat none in Hovorka-Mead et al. (2002) | The body that requires supervision, not "to protect integrity" |
| Who sees it | Ontario faulted a vendor's own use of recordings | Roles by name, not "authorised personnel" |
| How long it is kept | One page asked for deletion; one law allows no recording | A number of days, not "as long as necessary" |
| The alternative | Three of the five pages doubt consent where refusing costs the exam | A room, a date and a person, not "by enrolling you agree" |
| How to object | Ontario asked for an informal step before the formal one | A name and an address, not a link to the regulations |
Writing it, in order
- Decide exam by exam, and write down why a supervised room or a spoken check would not do. Where browser measures are enough, leave the camera off: EdProcto's integrity settings switch each measure on by itself.
- Ask students first where, and on what, they will sit the exam: it finds the shared rooms and weak connections early.
- Fix the alternative before the wording. A choice with nothing behind it is the consent three regulators doubted.
- Write the six statements in the present tense, with numbers: days, roles, dates.
- Publish it in one place, before enrolment closes and again a week ahead, and run a practice sitting with nothing at stake.
- Have your lawyer or data officer read it, and review it yearly: alternatives requested, recordings opened, objections made.
What does model consent wording look like?
Like the text below, which we wrote for this guide. It is a starting point for the institution's own lawyer or data officer, not a form to issue: it has been tried on no students and is claimed to satisfy no law. Every bracket is a decision.
Camera supervision for [exam], [date]
What is recorded. The picture from your webcam and the sound from your microphone, from the identity check until you hand in. [You will not be asked to show your room.] [No software analyses your face, eyes or voice.]
Why. This exam counts for [share] of [award], and [body] requires it to be supervised.
Who sees it. [Role] watches during the exam. Afterwards a recording is opened only by [roles], and only if [the invigilator noted something, or you ask]. [Company] stores it for us and may use it for nothing else.
How long it is kept. Recordings are deleted [number] days after marks are confirmed. One that is part of a review or an appeal is kept until that closes.
The alternative. You may sit the same exam at [place] on [date]. Tell [role] by [date]. No reason is asked, and your mark is unaffected.
If something is noted. A note is a question. [Role] will show you the moment and hear you before anything is decided.
How to object. Write to [data officer] at [address]. You may ask to see your own recording.
Your set-up. A practice session runs on [date]. If you lack a camera, a quiet room or a steady connection, tell [role]: [what is offered].
☐ I have read this page, and I choose [the camera sitting / the supervised room].
The last line asks for a choice between two real sittings. Where there is only one, call the page a notice and drop the box.
What does EdProcto record today, and what does its start screen say?
EdProcto is the best tool for an exam that needs set conditions, with a camera or without one, because each of its security measures is switched on by itself and its start screen states each one before the first question: the advance notice that the students in Hovorka-Mead et al. (2002) judged fairer.
Six act in the browser alone: "Leaving the tab", "Full screen", "Copy and paste off", "Right-click off", "Keyboard shortcuts off" and "Printing off". Each adds a line of text with its time to the attempt, and none records a picture or a sound.
Where any is on, a box headed "Monitoring" tells the student: "During this quiz your activity in the browser is logged. What is recorded is kept with your attempt for your instructor to review." Under "House rules" each measure stands beside its consequence, as in "Full screen: noticed and logged." The student ticks the same statement in the first person ("I understand that") to begin. Afterwards the student's result lists anything logged under "What this quiz logged", with the words "This is a record for your instructor to read. It is not a finding."
The camera measures belong to the Strict and Maximum levels: "Webcam on", "Microphone on", "A photo every few minutes", "Screen recording" and "A phone watching the room", and the face measures ("More than one face", "No face in view", "Eyes off the screen"), which are read in the student's browser. With any of them on, the same box names it, as in "During this quiz your camera is on, your microphone is on", and the photos are kept with the attempt for you to read. This guide applies to every one of them.
Free carries Light integrity, Tutor Medium and Educator Strict; Maximum comes with a licence arranged on a call. See pricing and for universities. No setting rules cheating out, a logged event is never proof, and every score is yours to release.
Quick questions
Can a university require a webcam for a remote exam?
This guide cannot say what your law allows. Ontario's report accepted online proctoring for certain exams and recommended an opt-out; Baden-Württemberg's law makes taking part voluntary.
What have regulators said about a room scan policy?
One German state's regulator says a camera pan through a student's room is not permitted under that state's law. Ask your data officer what applies to you.
How long should exam recordings be kept?
No study we found tested a period. One regulator recommended yearly deletion at the least; one court let stand rules deleting each recording after the exam unless something was noted.
Is a ticked box enough as proctoring consent?
Three of the five pages doubt it where refusing means not sitting the exam. A form offering a real alternative sitting stands on firmer ground; your lawyer decides.
Does EdProcto record picture or sound to monitor a quiz?
Only where you switch it on. Six of EdProcto's measures log browser events as text; its camera measures take photos for the attempt, its microphone measure listens for sound, and the start screen names each one before the student begins.
References
- Agencia Española de Protección de Datos. (2025, June 3). La AEPD sanciona el tratamiento de datos biométricos con IA en la evaluación universitaria online, aunque apunta posibilidades normativas [The AEPD sanctions biometric data processing with AI in online university assessment]. https://www.aepd.es/informes-y-resoluciones/criterios-juridicos-aepd/aepd-sanciona-tratamiento-datos-biometricos-ia
- Alessio, H. M., & Messinger, J. D. (2021). Faculty and student perceptions of academic integrity in technology-assisted learning and testing. Frontiers in Education, 6, Article 629220. https://www.frontiersin.org/journals/education/articles/10.3389/feduc.2021.629220/full
- Alge, B. J. (2001). Effects of computer surveillance on perceptions of privacy and procedural justice. Journal of Applied Psychology, 86(4), 797–804. https://pubmed.ncbi.nlm.nih.gov/11519663/
- Castelli, F. R., & Sarvary, M. A. (2021). Why students do not turn on their video cameras during online classes and an equitable and inclusive plan to encourage them to do so. Ecology and Evolution, 11(8), 3565–3576. https://pmc.ncbi.nlm.nih.gov/articles/PMC8057329/
- Duncan, A., & Joyner, D. (2022). On the necessity (or lack thereof) of digital proctoring: Drawbacks, perceptions, and alternatives. Journal of Computer Assisted Learning, 38(5), 1482–1496. https://doi.org/10.1111/jcal.12700
- Hovorka-Mead, A. D., Ross, W. H., Whipple, T., & Renchin, M. B. (2002). Watching the detectives: Seasonal student employee reactions to electronic monitoring with and without advance notification. Personnel Psychology, 55(2), 329–362. https://doi.org/10.1111/j.1744-6570.2002.tb00113.x
- Information and Privacy Commissioner of Ontario. (2024, February 28). Privacy complaint report PI21-00001: McMaster University. https://web.archive.org/web/20250430202507/https://decisions.ipc.on.ca/ipc-cipvp/privacy/en/item/521580/index.do
- Landesbeauftragter für den Datenschutz und die Informationsfreiheit Baden-Württemberg. (2021, July 17). Handreichung zu online-Prüfungen an Hochschulen [Guidance on online examinations at universities]. https://www.baden-wuerttemberg.datenschutz.de/handreichung-zu-online-pruefungen-an-hochschulen/
- Oberverwaltungsgericht für das Land Nordrhein-Westfalen. (2021, March 4). Eilantrag gegen videoüberwachte Prüfung der Fernuniversität Hagen erfolglos [Urgent application against video-supervised exam at FernUniversität Hagen fails] (14 B 278/21.NE). https://www.ovg.nrw.de/behoerde/presse/pressemitteilungen/01_archiv/2021/17_210304/index.php
- Persónuvernd. (2020, November 17). Uppfærðar leiðbeiningar um fjarkennslu [Updated guidance on distance teaching]. https://island.is/s/personuvernd/frett/uppfaerdar-leidbeiningar-um-fjarkennslu
- Rajab, M. H., & Soheib, M. (2021). Privacy concerns over the use of webcams in online medical education during the COVID-19 pandemic. Cureus, 13(2), Article e13536. https://pmc.ncbi.nlm.nih.gov/articles/PMC7996672/
- Whitelock, D., Edwards, C., & Okada, A. (2020). Can e-authentication raise the confidence of both students and teachers in qualifications granted through the e-assessment process? Journal of Learning for Development, 7(1), 46–60. https://jl4d.org/index.php/ejl4d/article/view/384