Interview Integrity Glossary
Key terms and definitions used in interview integrity monitoring, AI detection, and compliance reporting.
- AI Copilot
- Software that provides real-time AI-generated answers or suggestions to a candidate during a live interview, often designed to be invisible to interviewers and screen-sharing tools.
- Behavioral Anomaly
- A statistically significant deviation from expected human interaction patterns, such as unnatural gaze fixation, response latency spikes, or typing rhythm irregularities, that may indicate assisted responses.
- Candidate Integrity Score
- A composite risk score generated from multiple signal categories (behavioral, device, AI detection) that quantifies the likelihood of interview integrity violations during a monitored session.
- Data Handling Tier
- A configurable privacy level that determines what data is captured, retained, and displayed in integrity reports. Ranges from Standard (full capture) to Zero Retention (no persistent storage).
- Deepfake
- AI-generated synthetic media, typically video or audio, that replaces a real person's likeness with a fabricated identity. In interview fraud, deepfakes enable identity impersonation during live video calls.
- Device Fingerprint
- A composite identifier derived from system-level attributes (OS, browser, GPU, screen resolution, timezone) used to verify the authenticity and consistency of a candidate's interview environment.
- Evidence Package
- A structured collection of signals, metadata, and analysis results that constitutes the evidentiary basis of an integrity report, designed for audit review.
- Gaze Tracking
- Analysis of eye movement patterns during a video interview to detect indicators of off-screen reading, secondary device usage, or AI tool consultation.
- Integrity Report
- A structured document generated after a monitored interview session that summarizes detected signals, risk scoring, key findings, and recommended actions.
- Interview Fraud
- Any deliberate attempt to misrepresent a candidate's identity, knowledge, or capabilities during an interview process, including AI assistance, identity substitution, and proxy interviewing.
- Invisible Overlay
- A transparent or semi-transparent screen element that displays AI-generated content (answers, prompts, code) visible only to the candidate, not captured by screen-sharing software.
- Proxy Interview
- A scheme in which someone other than the actual candidate conducts the interview, either in-person with identity substitution or remotely via screen-sharing and deepfake tools.
- Remote Access Detection
- Identification of remote desktop software, screen-sharing tools, or virtual machine environments that may indicate a third party is controlling or viewing the candidate's session.
- Session Monitoring
- The continuous capture and analysis of system-level signals during a defined interview window, from session start to session end.
- Signal
- A discrete data point captured during a monitored session, such as a browser extension list, tab-switch event, or gaze deviation, that contributes to integrity analysis.
- Zero Retention
- The most restrictive data handling tier, in which no raw session data is stored after report generation. Only the final verdict and minimal metadata are retained.
Questions about interview integrity?
Our team can help you understand how these concepts apply to your organization.