Does HackerRank Use Your Camera? Webcam Rules by Mode

HackerRank uses your camera only when the employer has enabled Image Proctoring, Proctor Mode, or Desktop App Mode for that test. A default test and a Secure Mode test never request the webcam. When the camera is on, HackerRank captures still images at fixed intervals rather than continuous video, runs them through an image-analysis model, and shows the results to the recruiter. That is the documented behavior as of September 9, 2026.

Which HackerRank tests turn on the webcam

HackerRank's Test Integrity documentation lists three features that involve the camera.

Image Proctoring is a standalone add-on. It "Requires candidates to enable their webcam before starting the test. The system captures periodic snapshots during the test, which appear in the candidate's report for review." An employer can enable it on a test that is otherwise unproctored.

Photo Identification "Requires candidates to capture their image before starting the test." It is a single capture used to confirm who is sitting the test, not ongoing monitoring.

Proctor Mode and Desktop App Mode both include webcam monitoring as a required permission. Secure Mode, the tier below them, does not. The Secure Mode documentation lists a monitor check and full-screen enforcement but no camera.

Because a browser cannot open a webcam silently, the permission prompt on the setup screen is a reliable signal. No prompt means no camera for that test.

How often HackerRank captures webcam images

The interval depends on the mode, and HackerRank documents it in three places.

The Proctor Mode documentation describes the permission as "Webcam access: Allows the system to monitor and record the candidate through the webcam during the test," and states that the system "Captures webcam images every five seconds" for tests carrying candidate personal information. The Desktop App Mode documentation gives "webcam images every 15 seconds." The Image Proctoring documentation does not specify a number.

In every case the output is a series of snapshots. None of the documentation describes a continuous video recording of the candidate.

What the image analysis looks for

Snapshots feed a model HackerRank calls Image Analysis. The Image Proctoring and Image Analysis documentation lists four categories:

• "Multiple faces: Detects two or more individuals in the same frame"
• "Secondary faces: Detects two or more unique faces at different times"
• "No face: Detects when no face appears in the frame"
• "Static face: Detects static image projections"

The same page adds that "HackerRank supports Image spoofing detection that identifies candidates who attempt to simulate a live presence using static images or software."

Desktop App Mode extends this. Its documentation says "AI continuously monitors candidate presence. If the candidate is not visible or multiple faces are detected, a warning appears," and that the app flags "mobile phones and tablets in the candidate's webcam feed." It also describes gaze analysis of "eye and head movements" as a limited-availability feature where "Only high-confidence behavioral patterns are flagged as suspicious activity for review." What that gaze model can and cannot infer is examined in whether coding assessments track eye movement.

What the recruiter sees

Captured images land in the candidate's Summary Report. The Image Proctoring documentation says they appear under the "Attempt Activity" tab, where a reviewer can expand the "Proctor" section, and that flagged images are grouped by category with "up to five images" shown per category. In Proctor Mode and Desktop App Mode, webcam images are also synchronized with screenshots and integrity events in the Session Replay.

A flagged image is a model classification, not a finding. A housemate walking behind you produces a "Multiple faces" hit; stepping away to answer the door produces "No face." The report is built for a human to look at the frames, and whether that review happens is the employer's decision.

Consent and biometric data

The webcam permission itself is granted through the browser or, in Desktop App Mode, through the operating system. HackerRank's Candidate AI Notice states that "AI Features may use facial detection technology to compare images taken during the course of your assessment" and that "If HackerRank collects any biometric information, it will obtain your separate, informed written consent." If you are asked to sign a separate biometric consent, that is the trigger. How identity checks work across platforms, and how long they retain the data, is covered in the candidate guide to interview identity verification.

What this means for candidates

Check the setup screen. If it asks for the camera, HackerRank will photograph you every few seconds for the length of the test, and the model will flag any frame with no face or more than one. Sit alone, keep your face in frame, and keep phones out of view. The documentation states that "The test starts only if the candidate's face is clearly visible," so test your lighting before the timer. If it does not ask, the camera stays off for that test.

The screen-sharing half of the same permission screen is covered in when HackerRank records your screen. For the complete list of what each mode records, see what HackerRank proctoring detects and records.