When someone takes the Duolingo English Test at home, an AI system monitors the session by analyzing video, audio, typing patterns, and eye movements — looking for signs of rule violations across more than 75 different behavioral and environmental signals. AI flags potential issues, and then human proctors make the final call on whether to certify the test.

Each Duolingo English Test session is recorded through the test-taker's webcam, microphone, keyboard, and mouse. AI algorithms — using computer vision (analyzing what the camera sees) and natural language processing (analyzing response patterns) — review the session first, identifying potential violations. Human proctors then independently review flagged sessions and conduct their own checks. The full review process completes within 48 hours of test submission. Duolingo states that AI's role in proctoring is to assist human proctors rather than replace them. A peer-reviewed study on the system's validity and fairness was published in Educational Measurement, Issues and Practice (Wiley).