Strengthening exam security Combatting emerging threats with advanced ID verification

Author: Isabelle Gonthier, Chief Assessment Office, PSI

With technological advancements transforming the education and assessment landscape, the security of high-stakes exams has never been more critical. Assessment organisations must confront an evolving array of threats, from deepfake-assisted identity fraud to sophisticated digital forgeries. Conventional security measures are no longer sufficient to counteract these risks. Instead, an integrated, technology-driven approach to identity verification is essential to safeguard exam integrity and maintain trust.

This article explores the latest security challenges posed by proxy test takers, professional impersonators, and cheating rings. Sharing how assessment organisations can use artificial intelligence (AI) to strengthen their defences.

The rise of deepfake-assisted exam fraud

One of the most alarming developments in exam security is the increasing use of deepfake technology to facilitate proxy testing. Deepfakes use AI to create hyper-realistic images and videos that mimic a real test taker’s appearance and behaviour. With this technology, an impersonator can convincingly appear as a legitimate candidate during an online-proctored exam, making detection extremely challenging.

As deepfake sophistication grows, assessment organisations face a pressing challenge – how to differentiate between genuine candidates and AI-generated impostors. If left unchecked, deepfake-based fraud could undermine the credibility of qualifications and damage public confidence in assessment outcomes.

AI-powered deepfake detection

To counteract the threat of deepfakes, assessment organisations must deploy equally advanced AI-driven detection systems. These tools analyse subtle inconsistencies in digital images and video, identifying tell-tale signs of deepfake manipulation. Key detection techniques include:

• Facial anomaly analysis: AI scrutinises facial features for unnatural movements, inconsistencies in skin texture, and irregular blinking patterns that indicate digital tampering.
• Liveness detection: Advanced biometric authentication ensures that a live person is present by detecting micro-expressions and subtle physiological responses that deepfakes struggle to replicate.
• Real-time verification: Continuous monitoring during assessments helps flag potential identity mismatches, allowing human proctors to intervene when necessary.

By integrating deepfake detection into exam security protocols, assessment organisations can stay ahead of emerging fraud tactics and protect the integrity of online and remote assessments.

Digital document forgery: a growing concern

Another major risk facing assessment security is the rise of digital document forgery. In recent years, fraudulent digital ID documents have surpassed physical counterfeits as the leading method of identity fraud. The ease with which fake digital credentials can be created using AI tools makes them a preferred method for proxy test takers attempting to bypass ID verification checks.

According to PSI partner Entrust, digital forgeries accounted for 57% of all document fraud in 2024. This is a staggering 244% increase from 2023 and an astounding 1,600% surge since 2021, when nearly all fraudulent documents were physical counterfeits.[i]

National ID cards, driver’s licences, and even university-issued identification documents are vulnerable to digital manipulation. This growing threat necessitates the use of robust, multi-layered verification methods that go beyond traditional ID checks.

Strengthening ID verification with AI-driven biometrics

To combat digital ID fraud, AI-powered biometric verification offers a powerful solution. Key technologies include:

• Facial recognition cross-checking: AI compares the candidate’s real-time facial image with their official ID to verify their identity with precision.
• Keystroke dynamics analysis: Unique typing patterns, including speed and pressure, provide an added layer of security by verifying a candidate’s biometric signature.

These measures ensure that even the most convincing digital forgeries are detected, strengthening the credibility of exam and qualification systems.

Global Search: uncovering organised identity fraud

For large-scale assessment organizations, the threat of organised cheating rings and professional impersonators is a growing concern. Global Search technology addresses this issue by cross-referencing biometric data across multiple assessments to detect instances where the same individual appears under different identities. This sophisticated approach enables assessment organisations to:

• Identify repeat offenders attempting to manipulate the system.
• Prevent fraudulent candidates from using multiple identities to obtain unfair advantages.
• Strengthen exam security by maintaining an extensive biometric database to flag high-risk individuals.

By incorporating Global Search technology into assessment security frameworks, organisations can bolster their defences against increasingly sophisticated fraud tactics.

A multi-layered approach to exam security

Ensuring the integrity of UK assessments requires more than just adopting new technologies – it demands a holistic, multi-layered security strategy that combines AI-driven solutions with human expertise. By implementing deepfake detection, biometric authentication, and digital document verification, assessment bodies can stay ahead of the threats while maintaining a seamless test taker experience.

A robust security framework not only protects the credibility of qualifications but also upholds public trust. As threats continue to evolve, assessment organisations must remain proactive, continuously refining and enhancing their security measures to safeguard the future of testing.

For a deeper dive into AI-driven security strategies, download the PSI guide to AI and exam integrity

[i] Entrust Cybersecurity Institute, 2025 Identity Fraud Report.