Honorlock Explained: What It Detects, What It Doesn't, and What Students Should Know in 2026

What Honorlock Is, In One Paragraph

Honorlock is a hybrid AI-and-human proctoring service that monitors students during online exams. It runs as a browser extension on Chrome or Edge, activates when a proctored quiz starts inside Canvas, Blackboard, Moodle, D2L Brightspace, or another LMS, and records video, audio, and screen activity for the duration of the test. AI flags suspicious moments in real time. A human reviewer (either at Honorlock or your instructor) can review the session afterward.

What Gets Captured During a Session

Once a Honorlock session begins, several streams run in parallel.

• Webcam video. Continuous recording of your face and immediate surroundings.
• Microphone audio. Continuous recording of room audio.
• Screen activity. What you do inside the browser, including any new tabs or windows opened.
• Room scan. A 360-degree video of your environment, captured at the start of the session.
• ID verification. A photo of your student or government ID, captured at the start.
• Browser activity. The extension reports which browser windows are active and detects when you try to open another.

What the AI Actually Looks For

Honorlock's AI is trained on hundreds of thousands of hours of recorded exam sessions and is tuned to flag behaviors that correlate with cheating. Specific flags include:

1. Eye-gaze patterns. Looking away from the screen for long stretches, or looking down repeatedly (a signature of reading from a phone in your lap).
2. Multiple faces. Another person visible on camera, even briefly.
3. Voice events. Talking, whispering, or specific phrases like "Hey Siri," "Hey Google," or "Alexa," which the system listens for explicitly.
4. New browser windows or tabs. The extension reports the URL of any new window opened during the session.
5. Phone visibility. A phone visible on camera flags as a high-confidence event.
6. Camera obstruction. Covering the lens, leaving the frame, or moving the camera triggers a flag.
7. Idle and resume patterns. Unusually long pauses followed by rapid answering can flag, though this one has higher false-positive rates.
8. Background search detection. Honorlock seeds a small number of fake "answer" pages on the open internet for common test questions; if you search for and visit one, the system can detect that the search happened.

What Honorlock Cannot See

Honorlock is not all-seeing, and most students overestimate it. There are specific gaps.

It cannot directly see what is happening on a second device that is outside the webcam's field of view. It cannot read the content of a phone that never enters the frame. It cannot directly see what is written on a piece of paper next to you (though it can see you looking at it and flag the gaze pattern). It cannot read your mind, and it cannot prove what you were thinking when you looked away.

What it can do is flag the behavior. "Student looked down repeatedly for 12 seconds at minute 23" is the kind of flag that gets surfaced. Whether that flag becomes an academic integrity violation depends on the instructor, the institution's policy, and the broader pattern of the session.

How Flags Become Consequences

A flag is not a verdict. The chain of events between AI detection and any consequence looks roughly like this.

1. AI assigns an integrity score to your session based on the count and severity of flags.
2. A human reviewer (at Honorlock for high-stakes accounts, or your instructor for everything else) opens the session if the score is below a threshold.
3. The reviewer watches the flagged clips, not the full session.
4. If the reviewer believes a flag represents actual cheating, they note it and pass the case to the instructor.
5. The instructor decides whether to confront you, request more information, or refer the case to the school's academic integrity office.

Honorlock Versus the Other Major Proctoring Tools

Honorlock vs. Respondus Monitor: Honorlock relies more on AI and has lighter live human involvement. Respondus combines its LockDown Browser (which restricts your computer) with Monitor (which is the AI proctoring layer). If you have to install software rather than just a browser extension, you are probably using Respondus.

Honorlock vs. ProctorU: ProctorU traditionally uses live human proctors who watch sessions in real time, while Honorlock is mostly AI with humans reviewing afterward. Live proctoring is more invasive but also more able to intervene mid-test.

Honorlock vs. Proctorio: Proctorio is also extension-based and AI-driven, similar to Honorlock. The flag types are similar; the user interfaces differ.

Practical Tips Before a Honorlock Exam

These are not tricks. They are basic hygiene that prevents avoidable flags.

• Test the system days before the exam. Honorlock has a practice quiz at most schools. Run it. Find out now if your camera or microphone is failing.
• Set up the room. Plain background, good lighting, no second screen visible, no posters with notes, no piles of books open to relevant pages.
• Phone out of the room or face-down across the room. Not in your pocket, not on the desk.
• Tell people in your house. A roommate walking in or a parent calling your name is a flag.
• Close everything else before you start. Other apps and tabs get reported.
• Do not look down at your hands while you think. This is the single most common avoidable flag. Look at the screen or at the ceiling.
• Use a wired connection if you can. A network drop during the exam can corrupt the session and require a retake.

The Bigger Picture

Proctoring tools are imperfect. They miss cheating that happens off-camera, and they sometimes flag students who did nothing wrong. The right way to think about them is as a deterrent and a documentation system, not a perfect detector.

The reliable way to perform well on a proctored quiz is the same as the reliable way to perform well on any quiz: prepare for the format, know the material, and do not give the AI obvious flags to work with. If you have done the studying, the proctoring is mostly background noise.