How to protect that the person taking the test is the real candidate

Note: camera-based measures and behavioral signals may involve sensitive or biometric data. Their use requires the candidate’s consent and should be defined with your legal team according to each country’s regulations.

Protecting identity in a remote assessment means confirming, with the candidate’s consent, that the person taking the test is the one who applied and is answering on their own. It isn’t achieved with a single magic measure, but by combining signals —snapshots, IP-based location, behavior during the test— that provide reasonable confidence and assist a human review, never an automated verdict.

In an in-person process, the question of identity doesn’t come up: you see the person. In a remote one it’s a legitimate concern, because someone could ask another person to take the test for them. If that happens, the result no longer speaks about the candidate and your decision rests on a false foundation.

The signals that provide reasonable confidence

No signal on its own proves identity absolutely; together, they provide a solid foundation:

• Camera snapshots. Around eight captures per test, with consent, help verify it’s the same person throughout the process.
• IP-based location. A reference for where the test is taken helps detect inconsistencies, such as attempts to take it several times from disparate locations.
• Response time and latency. The pace of the answers provides context about how the test was taken.
• Facial behavior signals. Observations drawn from the snapshots that support human review, not a diagnosis or a disqualification.

The line that isn’t crossed

Protecting identity well respects the candidate; done badly, it intrudes on them. That’s why there are three non-negotiable safeguards:

1. Specific, informed consent. The candidate must know what measures are in place and what for before taking the test, not buried in fine print.
2. Proportionality. Use the measures necessary given the role’s risk, without turning every test into surveillance.
3. Human review. The signals assist a person who decides; never a system that disqualifies on its own.

Why this protects the honest candidate

It’s easy to think of identity verification as something done against the candidate. Done well, it’s the opposite: it protects those who answer honestly by ensuring they’ll be compared by the same standard as everyone else, not at a disadvantage against someone attempting to impersonate. Transparency —stating what measures are in place and what for— turns a measure that could feel invasive into a guarantee of fair play.

In short

Protecting that the person taking the test is the real candidate means confirming, with their consent, that they’re answering on their own. It relies on a set of signals —snapshots, location, behavior— that provide reasonable confidence and assist a human review, never an automated verdict. Because they touch on sensitive data, they require consent, proportionality, and a human decision, according to each country’s regulations. Done well, it protects those who answer honestly. Learn about the integrity controls or see how it works.