AI Proctoring Template: Enhancing Exam Security with Automated Monitoring

AI Proctoring Templates enable administrators to configure automated monitoring and security measures for online assessments. These settings leverage Artificial Intelligence to verify candidate identity, monitor exam environments, detect suspicious activities, and take automated actions when security policies are violated.

By using AI-powered proctoring, organizations can maintain exam integrity while reducing the need for continuous human intervention.

Auto Proctored Test / Assessment

Auto Proctoring uses Artificial Intelligence to continuously monitor candidates throughout the assessment and identify suspicious activities in real time.

The AI system analyzes candidate behavior, authentication data, and environmental conditions to detect potential violations and maintain exam security.

Proctoring Strictness

Administrators can choose a predefined proctoring sensitivity level based on the importance of the examination.

Mild Level

Best suited for mid-stakes examinations requiring standard monitoring and security controls.

Moderate Level

Designed for assessments that require enhanced monitoring and stricter violation tracking.

Strong Level

Recommended for high-stakes examinations where the highest level of monitoring and security enforcement is required.

Enabling a predefined proctoring level automatically configures multiple AI monitoring settings, escalation counts, and credibility score deductions, reducing manual configuration effort.

Note: If any individual AI configuration is modified manually, the predefined security level is automatically disabled.

Smart Proctor

Smart Proctor is an AI-powered monitoring assistant designed to emulate human supervision during online examinations.

The system continuously analyzes candidate activity and provides real-time notifications whenever suspicious behavior or unfair means are detected.

This helps candidates remain compliant with exam guidelines while allowing administrators to maintain a secure testing environment.

Candidate Authentication

Candidate Authentication verifies the identity of candidates before they begin the assessment.

The platform supports multiple authentication sources:

• Registration Photo

• Registration ID

• Capture ID

Registration Photo

Matches the candidate's live image against the photo uploaded on the Think Exam platform.

Registration ID

Matches the candidate's identity against the ID document uploaded on the Think Exam platform.

Capture ID

Captures the candidate's identity document in real time before the assessment begins.

Allow AI to Revoke Test

When enabled, the AI system automatically validates candidate identity using the configured authentication methods.

The AI compares images obtained from:

• Registration Photo

• Registration ID

• Capture ID

This multi-layer verification process significantly improves candidate authentication and reduces impersonation attempts.

Administrators can configure the maximum number of authentication attempts allowed.

If the candidate exceeds the configured attempt limit, the AI can automatically revoke the assessment and close the test window. The candidate must then restart the assessment process.

Continuity Camera

The Continuity Camera feature periodically verifies that the candidate remains present throughout the examination.

The system performs face verification at configurable intervals ranging from 10 to 30 minutes.

This feature also helps identify attempts to use static photographs instead of a live candidate by analyzing eye movement and facial activity.

Important: Continuity Camera does not generate Unfair Means (UFMs) and does not deduct credibility score.

Room Sanitization

Room Sanitization helps ensure that the candidate's examination environment is free from unauthorized materials.

The sanitization process may include:

• 360° Room Scan

• Desk Scan

• Body Scan

Administrators can configure the maximum number of sanitization attempts allowed.

AI-Based Object Detection

The AI system scans the environment during sanitization and identifies unauthorized or suspicious objects.

This helps prevent candidates from accessing prohibited materials during the examination.

AI Revocation During Sanitization

When Allow AI to Revoke Test is enabled, the AI can automatically revoke the assessment if room sanitization requirements are not met within the permitted number of attempts.

This ensures that candidates enter the examination only after successfully completing the required security checks.

Additional Camera Monitoring

AI Proctoring supports the use of an additional camera for enhanced monitoring.

Available camera views include:

• Side View

• Front View

• Back View

• Custom View

For Custom View, administrators can provide specific instructions regarding camera placement and angle.

Additional camera feeds help strengthen monitoring coverage and provide greater visibility during the examination.

Note: Face Not Present and Face Mismatch violations are not tracked through additional cameras.

Credibility Score Management

Each candidate begins the examination with a credibility score of 100.

The score decreases whenever AI-detected violations occur during the assessment.

Administrators can configure thresholds that determine how the platform responds to suspicious behavior.

Suspicion Threshold

When a candidate's credibility score falls below this value, the session is flagged as suspicious and may require further review.

Failure Threshold

When a candidate's credibility score falls below this value, the session is considered compromised.

AI-Detected Unfair Means (UFMs)

The AI Proctoring system can detect various types of unfair means during an assessment.

The supported UFMs include:

• Face Mismatch

• Face Not Present

• Looking Away

• Multi Face Detected

• Object Detected

• Permission Revoked

• Screen Focus Lost

• Voice Detected

Each UFM can be configured with an Escalation Count.

The Escalation Count determines when credibility score deductions begin.

Example

If the Escalation Count for Face Mismatch is set to 1, the first occurrence will only be recorded. Credibility score deduction begins from the second occurrence onward.

Automated Session Actions

AI Proctoring can automatically take actions based on credibility score thresholds.

Suspend Session

The candidate's session is temporarily paused when predefined conditions are met.

Terminate Session

The candidate's session is permanently ended when serious security violations are detected.

AI-Based Penalty Enforcement

When the Penalty setting is enabled, the AI system can automatically:

• Suspend candidate sessions

• Terminate candidate sessions

These actions are triggered based on configured credibility score thresholds and detected violations, helping administrators enforce exam security policies automatically.

Benefits of AI Proctoring Templates

AI Proctoring Templates provide a comprehensive framework for maintaining exam integrity through automated monitoring and enforcement.

Key benefits include:

• Automated candidate identity verification

• Real-time monitoring and violation detection

• AI-powered room sanitization checks

• Continuous candidate presence verification

• Credibility score-based risk assessment

• Automated session suspension and termination

• Reduced dependence on manual invigilation

• Improved security for high-stakes examinations

By configuring AI Proctoring Templates effectively, administrators can conduct secure, scalable, and reliable online assessments while minimizing operational overhead.

For complete template based feature explaination: Click Here