AI-assisted monitoring has transformed remote testing, but physical test centres have seen far less change—even as fraud becomes more sophisticated. Assessment organisations now face a difficult question: how do you strengthen security without creating unnecessary friction for test takers?

This session examines how AI-driven detection, real-time centralised oversight, and human judgement can work together in physical testing environments. Drawing on the development and early implementation of ProGuardian, Prometric will share practical lessons on minimising false positives, designing human-review and escalation workflows, integrating monitoring into test centre operations, and addressing privacy, governance, and regulatory expectations relevant to European assessment organisations.

Attendees will explore how security data can support more consistent decision-making and how effectiveness can be measured beyond detection accuracy, including response speed, unnecessary escalations, operational consistency, and test-taker experience. These lessons provide a practical framework for evaluating how emerging security technologies can strengthen exam integrity and protect credential value while preserving fairness, transparency, and trust—supporting the broader goal of Building Assessment for Good.
KEY LEARNING OBJECTIVES
By the end of the session, attendees will be able to:
• Explain how AI-driven detection, real-time oversight, and human judgement can work together to strengthen security in physical testing environments.
• Identify approaches for strengthening fraud detection while minimising false positives and unnecessary disruption for test takers.
• Evaluate the operational, governance, privacy, and performance considerations involved in introducing AI-assisted security into test centre environments.