Can Online Tests Detect Cheating? What Every Tool Actually Catches

What each tool records (quick comparison)

The short version: tools that watch the test (proctoring and lockdown browsers) and tools that check the work afterward (plagiarism and AI scanners) do most of the work. What the rest record depends on how the instructor sets things up.

The three ways online tests detect cheating

Every tool on this page falls into one of three groups, and most schools use more than one at a time.

• During the test: supervision. Proctoring software, lockdown browsers and device monitors watch the session itself: the webcam, the microphone, the screen, and whether other apps are open. This is the most direct kind of detection.
• From the platform: activity data. Learning platforms record when you opened a quiz, how long each question took, when you submitted, and sometimes which device or network you used. Unusual patterns stand out in these logs, and instructors can review them alongside the answers themselves.
• After the test: checking the work. Plagiarism scanners compare written answers against papers, web pages and other students’ submissions, and most now include an AI-writing indicator. Instructors also compare answers across a class by eye.

A 2022 systematic review of online-exam cheating research (in the PMC archive) describes the same shift toward these automated checks and is a readable overview of what the research covers.

How learning platforms (LMS) detect cheating

A learning management system like Canvas, Blackboard, Moodle or Google Classroom records far more than students expect: when you opened the quiz, how long each question took, and, with a proctoring or lockdown add-on, whether you left the test window. On their own they catch less, which is why most schools pair them with a plagiarism or proctoring tool.

The detail by platform: how Canvas detects cheating, what professors can see on Canvas, whether they can see your tabs, Grammarly on Canvas, ChatGPT on Canvas, Blackboard, Moodle, Schoology, Google Classroom, Google Forms and ALEKS.

How remote proctoring tools work

Proctoring software is the most direct form of detection. It uses your webcam, microphone and screen, and many products add AI that flags behavior such as looking away for long periods, a second face, another voice, or a phone in view. Some also search the web for leaked exam questions. Live proctoring adds a person watching in real time; automated proctoring records the session and flags moments for later review.

One point students often miss: a flag is not a verdict. It marks a moment for the instructor or school to look at, and the decision about whether anything was wrong is made by a person, usually after reviewing the recording.

Read how Proctorio, ProctorU and Honorlock work, what LockDown Browser records, whether Examplify records you, whether GoGuardian can see your screen, and what happens after a ProctorU incident report.

Homework, answer sites and AI chatbots

Sites like Chegg, Course Hero, Quizlet, Mathway and Photomath are not cheating by default, but using them to complete graded work usually is, and schools can often trace it. Chegg gives verified school officials the dates and times a question and its answer were posted during academic-integrity cases, and Course Hero terminates accounts found to have broken a school’s policy. AI chatbots raise the same question in a new form: whether using one is allowed depends on your instructor’s rules for that assignment.

The detail: Chegg and whether it notifies your school, Course Hero and whether it notifies your school, Quizlet, Mathway, Photomath, and whether using ChatGPT counts as cheating.

Plagiarism and AI-writing checkers

Plagiarism scanners compare your text against a large database of papers, web pages and other students’ submissions, and most now flag likely AI-generated writing too. They report a similarity score or an AI estimate, and an instructor decides what that number means in context.

AI detection deserves extra care because it can be wrong. Turnitin, the most widely used checker, says in its AI writing detection FAQ that the AI percentage “should not be used as the sole basis for action or a definitive grading measure by instructors.” It also says it aims to keep false positives under 1% for documents with more than 20% AI writing, which means some human-written work will still be flagged. If that happens to you, your drafts, notes and document version history are the best evidence you have.

More detail: whether Turnitin detects ChatGPT, how Turnitin handles QuillBot, what a Turnitin similarity score means, how SafeAssign checks homework-site content, whether SafeAssign detects QuillBot, the best plagiarism checkers for teachers, and whether checkers store your paper.

How detection differs: school, college and professional exams

The tools a student meets depend heavily on where they are studying. In K-12 schools, most monitoring happens on school-managed devices and accounts: Chromebooks running a device monitor like GoGuardian, Google Forms quizzes in locked mode, and curriculum platforms such as Savvas Realize that report time on task. Teachers usually see this in real time during class, and the follow-up is a conversation rather than a formal hearing.

In college, the emphasis shifts to the learning management system plus add-ons. A typical online midterm might run in Canvas or Blackboard with a lockdown browser, and written work goes through Turnitin or SafeAssign. Fully online programs and large classes are the most likely to use automated or live proctoring, and suspected violations go through the school’s formal academic-integrity process.

Professional and licensing exams are the strictest. Secure exam apps such as Examplify lock the device, and many certification exams use live remote proctors or require a test center. The stakes are higher too: a violation can mean a cancelled score or a ban from retesting, set out in the candidate agreement you accept before the exam.

For teachers: setting up an online test that is harder to cheat on

Detection works best alongside a test that is designed with integrity in mind. Most of the options below are built into common platforms and cost nothing extra:

• Draw questions from a pool and shuffle them, so neighboring students see different questions in a different order.
• Set a time limit that fits the test, long enough to think, short enough to discourage looking things up for every question.
• Ask applied questions, such as explaining a method or working with a new example, which are harder to copy from an answer site.
• State the rules for AI tools and outside sources in the test instructions, so students know exactly what is allowed.
• Use a lockdown browser or proctoring only where the stakes justify it, and tell students in advance what it records.
• Run written answers through a plagiarism checker, and treat any score, especially an AI score, as a reason to look closer rather than as proof.

Communication matters as much as the settings. Clear rules remove the guesswork that leads to honest mistakes, so spell out in the syllabus and the test instructions which resources are allowed, whether calculators or notes are permitted, how AI tools may or may not be used, and what happens if a check raises a question. It also makes any later conversation about a flag fairer for everyone.

For the tools themselves, see our comparison of plagiarism checkers for teachers and how test banks affect question security.

Privacy: what these tools record about you

Proctoring and monitoring tools collect personal data, so it is reasonable to ask what they keep. Depending on the product and settings, that can include webcam video, audio, screen recordings, a scan of your room, and a photo of your ID. Device monitors such as GoGuardian work on school-managed devices and accounts.

Who can see the recordings, and how long they are kept, depends on both the vendor and your school’s contract with it. Before a proctored exam, read your school’s guidance and the vendor’s privacy policy, and ask your instructor if anything is unclear. If you need an accommodation, such as a quiet-room exception or a different testing setup, ask for it before the exam rather than after.

What happens if you’re accused

Detection is only the first step. What follows is an academic-integrity process that varies by school but usually includes a written notice, a chance to respond or meet, a decision, and a right to appeal. Responding calmly and in writing, asking what evidence was used, and keeping your drafts and notes all help.

More help: what to do if you are accused of plagiarism, how to respond when a teacher accuses you of cheating, how to email a professor, whether using old exams counts, test banks, and how much professors actually care.

Frequently asked questions

Can online tests detect cheating?

Yes. Online tests detect cheating in three main ways: proctoring software that watches the session (webcam, microphone and screen), platform logs that record timing and activity, and plagiarism or AI-writing checks run on the work afterward. How much a particular test records depends on which of these the instructor turns on.

Can teachers see if you open other tabs during an online test?

It depends on how the test is set up. Proctoring tools and lockdown browsers can detect or block tab-switching, school-device monitors like GoGuardian show the screen directly, and the instructor’s settings decide which of these apply.

Does Turnitin detect ChatGPT?

Turnitin has an AI-writing indicator that estimates how much of a submission was likely generated by tools like ChatGPT, alongside its standard similarity check. Turnitin itself says the AI percentage should not be used as the sole basis for action, and that it aims to keep false positives under 1% for documents with more than 20% AI writing.

Is using Chegg or Course Hero cheating?

Using them to study is not, but submitting their answers as your own work usually breaks academic-integrity rules. Chegg gives verified school officials the dates and times a question and its answer were posted, while saying it does not share personal information, and Course Hero terminates accounts found to have broken a school’s integrity policy.

Can Canvas or Blackboard detect cheating on their own?

Only partly. Both log quiz activity, but most detection comes from the add-ons schools layer on top, such as Turnitin, SafeAssign, Respondus LockDown Browser or a proctoring service.

Does a proctoring flag mean you cheated?

No. A flag marks a moment for someone to review, such as looking away or a second voice. The instructor or school decides what it means, usually after looking at the recording and the rest of the evidence.

Can AI detectors be wrong?

Yes. No AI-writing detector is perfect, which is why Turnitin tells instructors not to use its AI score as the sole basis for action. If you are flagged for writing you did yourself, drafts, notes and version history are the strongest evidence you can bring.

What should I do if I’m accused of cheating on an online test?

Read the school’s academic-integrity policy, respond to the notice calmly and in writing, ask what evidence was used, and keep anything that shows how you did the work. Most schools have a meeting or hearing step and an appeal.

The bottom line

Online cheating is detectable, but not by magic. It comes down to three things: tools that watch the test (proctoring and lockdown browsers), tools that check the work (plagiarism and AI scanners), and the data platforms quietly log. None of them decides on its own whether someone cheated; a person reviews what they find. For students, the safest assumption is that at least one of these is switched on. For teachers, the strongest setup combines good test design with one or two of these tools, used transparently.