Recruitment security has started to overlap with test security because proxy interviews, AI-assisted cheating, and deepfake-enabled impersonation all attack the same basic question: is the person in front of us really the person we think they are? The source set suggests that employers are increasingly returning to in-person interviews and stronger identity checks when the risk is high, but that does not settle the broader policy question of how much friction is proportionate. The strongest reading is that recruitment programmes need the same layered thinking used in assessment security: identity assurance, task design, and review routes that match the risk.

This comparison looks at how recruitment fraud patterns map to test-security controls. The source note discusses proxy interviews, test-taker impersonation, AI cheating, and deepfakes, alongside employer responses such as in-person interviews and AI proctoring tools.

If a recruitment process can be bypassed by someone else answering, speaking, or appearing on behalf of the candidate, then the whole talent decision becomes less trustworthy. For sectors that rely on licensure, technical competence, or public confidence, that is the recruitment equivalent of exam impersonation.

- **Proxy interview**: another person attends or answers for the candidate. - **Deepfake**: synthetic audio or video used to imitate a person. - **Identity assurance**: the checks used to confirm who is actually present. - **Friction**: the time, inconvenience, and cost added by stronger controls.

The source suggests that in-person interviews still have a role when identity risk is high, because they make proxy participation harder. It also fits the broader security pattern that visual and behavioural cues alone are not enough when synthetic media is plausible.

The contested issue is how much of recruitment should move back to physical settings. Some employers will accept more friction to gain confidence, while others will prioritise scale and accessibility and rely more on layered digital checks. Another open question is whether AI proctoring in recruitment is a durable fix or just a stopgap. As with testing, the tool can help, but it does not remove the need for sound process design and human review.

- onboarding or hiring the wrong person - paying for digital controls that do not stop deepfake impersonation - excluding candidates through unnecessary friction - confusing detection with proof

1. Identify which parts of recruitment are most exposed to identity fraud. 2. Decide whether in-person review is justified for the highest-risk stages. 3. Check whether the process needs a deeper identity step before interviews or assessments. 4. Use AI or digital checks as layers, not as the whole control model.

| Approach | Strength | Limitation | Best fit | |---|---|---|---| | **In-person interviews** | Stronger identity assurance | Slower and less scalable | High-risk or high-stakes hiring | | **Digital interview with identity checks** | Flexible and more scalable | More exposed to impersonation or deepfakes | Routine hiring with moderate risk | | **AI-assisted screening** | Efficient triage | Needs careful governance and human review | Large applicant volumes |

A hiring team notices increasing concern about candidates using proxy interviews. It moves the final interview round back on-site for critical roles and keeps a lighter digital screen earlier in the process. That gives the organisation a stronger final identity check without making every stage equally burdensome.

- Recruitment-fraud source note discussing proxy interviews, AI cheating, and deepfakes.

The relevant market here includes recruitment platforms, identity-check suppliers, and AI proctoring providers. Their claims often sound similar to test-security vendors, which is exactly why buyers should separate identity assurance from simple surveillance.

### Is this really a test-security topic? Yes, because it uses the same integrity logic: the decision is only trustworthy if the right person is being assessed. ### Are deepfakes a real hiring risk? Yes. They can be used to misrepresent identity in remote processes if controls are weak. ### Do I need in-person interviews for everything? Not usually. The question is whether the role and risk justify that extra control at the most sensitive stage.

Tim Burnett (Admin)

`Test Community Network. "Proxy interviews, AI cheating, and deepfakes in recruitment security." TCN Wiki. Last reviewed 2026-05-05. https://www.testcommunity.network/wiki/test-security-comparison-proxy-interviews-ai-cheating-deepfakes`

- Recruitment-fraud source note discussing proxy interviews, AI cheating, and deepfakes.