AI interview proctoring
Integrity rules that apply while the interview runs.
Many tools only flag cheating after the interview ends. Firstview applies the rules during the interview and pauses it when one breaks.
The proctoring rules
| Rule | What Firstview checks |
|---|---|
| 1Whole screen shared | The candidate shares the entire screen, not one window. The check also runs on the server, so the browser cannot fake it. |
| 2Full screen | The interview stays in full screen. |
| 3Camera on | The camera stays on. Virtual cameras are blocked and pause the interview. |
| 4Enough light | The camera image stays bright enough to see the candidate. |
| 5On the interview tab | Leaving the tab for 2 seconds or the window for 10 seconds counts as a break. |
| 6One screen | The interview does not start while a second screen is connected. A screen connected later is noted in the AI interview report. |
A fair way back
Wi-Fi drops and closed laptop lids happen to honest candidates too. When a rule breaks, the interview pauses and the candidate gets 4 minutes to fix it. After a dropped connection, they get 5 minutes to rejoin. If they do not come back, the interview ends.
- Rule breaks
- Interview pauses
- 4 minutes to fix it
- Continue
- or the interview ends
What Firstview does not do
Firstview does not analyze faces, expressions or the content of the candidate’s screen, and video is never part of the evaluation. The camera checks are only about light and the camera source.
Limits, stated plainly
On a phone, the screen-share and full-screen checks cannot run, so the interview continues with fewer checks. Firefox is not supported, because it cannot report which screen is shared.
Questions about Interview Integrity
How do you stop candidates cheating with ChatGPT in an online interview?
Firstview makes it hard to use ChatGPT on the interview device. The candidate shares the whole screen and stays in full screen, and leaving the interview tab for 2 seconds or the window for 10 seconds pauses the interview. It cannot see a second device such as a phone, so it reduces cheating but does not prove that none happened.
Does Firstview verify identity?
No. It checks the interview conditions, not ID documents.
Is this fair to candidates?
Every candidate gets the same rules and the same 4 minutes to fix a break.
Why does interview integrity matter now?
In a 2025 Gartner survey of 3,000 job candidates, 6% admitted to interview fraud, such as having someone else pose as them. Gartner predicts that by 2028, one in four candidate profiles worldwide will be fake.
Is this AI interview proctoring?
Yes. Firstview proctors the AI interview live with six rules, and it doesn’t analyze faces or expressions.
Does it stop deepfakes?
It blocks virtual cameras, a common way to feed a fake video into an interview. It doesn’t analyze faces, so it isn’t a deepfake detector.
Works with the rest of Firstview
Interview Integrity runs inside the AI interview in Firstview. The AI interview report that follows sits next to the CV evaluation on the candidate record, and the AI Assistant can answer questions about both.
