Integrity controls
Integrity controls
What integrity controls are
When a candidate takes an assessment online, the team needs confidence that the result reflects their own performance. Integrity controls are context signals recorded during the session —always with consent— that the team can review alongside the rest of the evidence. They are not a detector or a verdict: they are comparable context to support human reading.
They work in two complementary layers: design measures that shape the assessment in advance, and in-session monitors that record environment signals while the candidate answers.
Layer 1: design measures
The first layer acts before the session: it is built into how the assessment is designed. The goal is for every candidate to take it under comparable conditions, with no need to watch over anyone.
- Large question banks and randomized order, so two people don’t see exactly the same thing.
- Calibrated time limits per assessment, keeping the same yardstick for everyone.
- Unique invitation links per candidate.
- Equivalent versions of the same test for recurring processes.
Layer 2: in-session monitors
The second layer records context signals while the candidate answers. Each of these signals is optional, configured by role and country, and only activated with the candidate’s informed consent. They are presented to the team as context, not as an honesty score.
- Camera and environment: with consent, records context signals from the assessment environment.
- Full screen and window changes: records whether the assessment loses focus, as a context signal.
- IP and device: helps confirm session continuity and technical conditions.
- Pace and response times: add comparable context about how the test was taken.
None of these signals issues a conclusion on its own. They are delivered to the recruiter to interpret alongside the rest of the evidence and to decide, with human judgment, whether a point is worth discussing with the person.
What they do NOT do
- They don’t issue a cheating or honesty verdict.
- They don’t automatically disqualify any candidate.
- They don’t watch or record anything without consent.
- They don’t replace the recruiter’s judgment or the interview.
- They don’t decide for the team: the final decision is always human.
Frequently asked questions
Do integrity controls detect whether someone is cheating?
They don’t work that way. The controls record context signals during the assessment —with consent— and show them to the team. They don’t issue a cheating or dishonesty verdict: they add context so a person can review and interpret it.
Does a signal automatically disqualify a candidate?
No. No signal disqualifies on its own. Signals support human review; the team interprets them alongside the rest of the evidence and decides whether to discuss the point with the person.
Is the camera used without the candidate knowing?
No. Any sensitive signal —camera, IP, full screen— requires the candidate’s informed consent and is disclosed before the assessment begins. The candidate knows what is recorded and why.
What if the candidate doesn’t want to turn on the camera?
Consent belongs to the candidate. If they decline an optional signal, the team knows and takes it into account during review; it is not an automatic reason for exclusion. How to proceed is a human decision.
Do integrity signals replace the recruiter’s judgment?
No. They are one more layer of context, not a conclusion. The recruiter still reads the full body of evidence, weighs it and decides. The controls help prioritize what to look at more closely.
What signals are recorded during the session?
Depending on the configuration and consent, context signals such as window or full-screen changes, response pace or environment signals may be recorded. They are presented as comparable context, not as an honesty score.