BrightHire Interview Integrity helps teams identify potential candidate fraud directly within the interview workflow. It surfaces reviewable signals across BrightHire Screen AI interviews and live interviews, giving recruiting teams more context when something may need a closer look.
​

Interview Integrity is available to all BrightHire customers. It supports human review and does not automatically reject candidates or make hiring decisions.
​

Signals are grouped into the following configurable categories. Deepfake Detection can be used on its own, and Session Enrichment requires Session Metadata. Which ones you can use depends on your interview platform and setup.

BrightHire brings these signals into one place so your team can review the evidence and decide what action, if any, is appropriate.

Available signals depend on the interview type and your configuration.

Behavior signals are only available in BrightHire Screen, because BrightHire runs that interview experience in the browser. For live interviews, Deepfake Detection is available for customers using Greenhouse, Ashby or Workday, where candidate consent can be collected. See Candidate Consent below.

Once Interview Integrity is configured, eligible new interviews are analyzed automatically. Interviewers do not need to change anything about how they run interviews.

After an interview, BrightHire analyzes the available interview data and shows any integrity signals in the interview's Signals tab. Interviewers do not see live warnings during the conversation, so a signal cannot bias them while they are still talking with the candidate.

Each signal comes with supporting evidence and, where relevant, links to the moments in the recording. It is an input for review, not a final determination.

Integrity analysis does not delay normal interview processing. If analysis takes longer or fails, the recording, transcript and notes still process as usual.

Interview Integrity does not:

Deepfake Detection always requires explicit opt-in consent from the candidate. If consent is missing, declined or cannot be matched to the candidate's application, BrightHire does not run Deepfake Detection for that interview.

For BrightHire Screen AI interviews, candidates are asked for consent within the Screen interview itself.

For live interviews, candidates give consent through your ATS application, and BrightHire picks it up automatically on the next ATS sync. Consent collection is supported for Greenhouse, Ashby and Workday:

Each ATS article includes sample consent wording as a starting point. It is not legal advice, so review it with your legal team before using it.

Deepfake consent is separate from interview recording consent. Your existing recording consent process does not change, and a recording consent answer is never used as deepfake consent.

Deepfake Detection is the only signal that uses biometric data. The other signals (Behaviors, Session Metadata and Session Enrichment) do not use this consent step, and BrightHire does not analyze facial expressions for them. Review your candidate notices with your legal or compliance team before rollout.

Candidates do not need to install anything for Interview Integrity to work.
​
If you use Deepfake Detection for live interviews, candidates see one extra consent question in your application. If a candidate declines, BrightHire does not block live interviews; Deepfake Detection simply does not run. How you handle a declined consent is your organization's decision.
​
For BrightHire Screen AI interviews, candidates are asked for consent within the Screen interview itself and declining the consent does block the interview if Screen Deepfake detection is on. The candidate must check the consent box to proceed.

Your BrightHire CSM or Support will walk you through setup. What you need depends on the signals and interview platforms you want to use.

Behavior signals in BrightHire Screen need none of these steps. Google Meet needs no app install.

For Zoom, your Zoom admin installs or reinstalls the BrightHire Notetaker app and grants the View meeting participants' metrics permission. Deepfake Detection on Zoom does not need the app. See the Zoom setup article for details, including how this works with your recording disclaimer settings.

Once the requirements are in place, a BrightHire Site Admin turns on the signals:

Admins can limit who sees Interview Integrity signals by organization role and extend access to specific hiring teams. Users without signal access can still open the interview as usual; they just don't see the Signals tab.

Go to Settings → Roles and Policies and set:

Site Admins always have access.

Adjust these to match who is responsible for reviewing signals in your process, such as TA, security, or specific hiring managers.

Signals appear in the Signals tab of a completed interview, in the Signals column of the interview list, and on the candidate page. If you use the BrightHire Chrome extension, you can also see them in the side panel in your ATS.

Each interview shows an overall risk level. It reflects how strong the detected evidence is, not the probability that fraud occurred.

The overall level matches the highest-risk signal category. For example, Moderate network risk plus a Strong deepfake signal shows as Strong overall.

The review outcome and note are visible to your team, so others know the signals have been looked at. Recording an outcome does not clear signals or change the candidate's status.
​