CoderPad detects cheating by recording how code arrives in the editor and flagging the events that break the pattern. As of September 9, 2026, its documentation describes four signals: a paste from outside the pad, the browser leaving the pad, a submission identical to another candidate's, and a question finished implausibly fast. CoderPad Interview surfaces the first two to the interviewer live and again in playback. CoderPad Screen, the unattended assessment, adds full-screen enforcement, optional webcam snapshots, geolocation checks, and automatic plagiarism matching. In both products CoderPad tells reviewers the same thing: "Treat alerts as prompts for clarification, not immediate disqualification."
The two signals every pad records
CoderPad's cheating prevention documentation for Interview states that "CoderPad alerts interviewers when a candidate: Pastes code from an external source" and "Leaves the IDE (e.g., tab-switching)." Those alerts appear during the interview and persist in the replay.
The playback documentation describes how they look. A yellow dot marks each paste from outside the pad with a "Code pasted" tooltip. An orange dot marks each loss of focus with a duration, for example "Clicked away for 19 seconds." A separate alert fires when a candidate copies all the code in a file. The pad summary collects these into a suspicious-behavior section next to talk time and code runs.
The underlying record is broader than the alerts. "Every pad records a complete timeline of the session, including: Code edits, Runs, Cursor movements, Copy/paste events, IDE exit notifications." CoderPad's guidance to reviewers is to "Look for bursts of activity inconsistent with normal typing." That is the signal a paste-free workaround still produces: code that appears in the editor faster or more finished than it was thought through.
How a browser sees a tab switch, and why it cannot see where you went, is explained in how interview websites detect tab switching.
What Screen adds when no one is watching
Screen runs unattended, so CoderPad layers controls on top of the recording. Its cheating prevention documentation for Screen, updated April 17, 2026, lists them.
Full screen and monitors. "Candidates must enter full-screen mode before starting the test. If they exit full-screen mode or switch to another monitor, an alert will trigger after a 10-second grace period."
Plagiarism. Screen will "Recognize if a candidate submits the exact code previously submitted by another candidate." The documentation describes an exact match, not a similarity score.
Speed. Screen flags questions completed unusually quickly relative to the expected time.
Copy and paste. Recruiters can "Automatically block copy/paste of the question statement" and can block pasting of externally sourced text into the editor. Where paste is allowed, the Screen code playback marks it "Red when your candidate pastes code."
Randomized questions. Question blocks can be randomized so "no two tests are the same," which blunts leaked answer keys.
Geolocation. Screen records location changes during a test.
Webcam. On enterprise plans, optional snapshots are analyzed for no face, multiple faces, or a different face. "Candidates are not blocked from taking the test even if their webcam is inactive," per the webcam proctoring documentation.
Screen also tells candidates it is doing this. The documentation describes welcome screens and confirmation pop-ups that warn against cheating, and a tutorial that shows the report with code playback.
What the reviewer sees
In Interview, the interviewer sees alerts as they happen and CoderPad's instruction is to "Ask the candidate why they pasted code or left the IDE when alerts appear." Afterward the replay shows the whole session with the dots in place.
In Screen, recruiters see red exclamation marks in the candidate list, plagiarism alerts in the report, and warnings at the question level. Screen can generate follow-up questions with ChatGPT that ask the candidate to explain the logic of the code they submitted, and recruiters can override automatic scores. What a flag in an assessment report actually does to an application, across vendors, is covered in what happens when a coding assessment flags you.
AI is handled by making it visible
CoderPad's approach to AI is not to hunt for it but to bring it inside. Interviewers are told to "Enable AI Assist for candidates so all AI usage is visible within the platform," and candidates are told that "Any AI usage should remain within the AI Assist tab." The AI Assist documentation confirms the record: "All pad users' prompts and AI output will be saved for review in playback mode." Screen does the same for project questions.
That design has a consequence. AI used outside the pad leaves the same trace as any other outside source: a paste, a focus loss, or a burst of finished code. AI used inside it leaves a prompt log. Candidates are expected to "explain, critique, and adapt AI-generated output," which is the interviewer's real test either way.
What CoderPad does not document
No keystroke logging outside the editor. No screen capture. No stored video of the interview call. No similarity scoring in Interview, only in Screen, and there only as an exact match. No automatic rejection in either product. Absence from the documentation is not proof of absence, but it is the boundary CoderPad has chosen to publish, and the boundary that matches how a browser page works. What the pad can and cannot see of your display is in whether CoderPad records your screen.
What this means for candidates
Type in the pad. A paste is a marker, a tab switch is a marker, and a finished block after either is the thing CoderPad tells interviewers to ask about. If you need a reference, ask whether it is allowed and, if it is, say what you are opening. If AI Assist is on, use it there and be ready to defend every line. If it is off, that is the rule, and preparing with AI before the interview is the way to use it. For the format and a test-day checklist, see the CoderPad interview format and preparation guide.
Share
Explore with AI
Continue exploring
Related guides
Assessment Guides · 5 min read
Can Codility Detect ChatGPT? Typing Patterns and AI Matches
Codility names ChatGPT and Claude in its Typing Pattern Detection and matches submissions against AI-generated solutions. What each layer flags and its limits.
Assessment Guides · 5 min read
Does HackerEarth Detect Cheating? What It Flags
HackerEarth flags code matching another candidate above 85%, runs AI checks against public sites, and scores an integrity index. What each flag proves.
Assessment Guides · 5 min read
Does Mettl Detect Cheating? Flags, Proctors, and the Index
Mercer Mettl rolls browser rules, eight AI flags, optional live proctors, and a code similarity check into a credibility index. What each catches, what a flag means.