AI Policy

Last updated: [11/2025]

At Certify Assessment Solutions (CAS), we use Artificial Intelligence (AI) responsibly to support secure, fair and efficient assessment delivery. This page explains how AI is used within our platforms, what it does, what it does not do, and how we ensure human oversight at all times.

1. What AI We Use

CAS uses AI-supported features within:

SmarTest Invigilate

Our remote proctoring platform uses AI to:

• Detect potential anomalies (e.g., unusual movements, additional faces, suspicious audio)

• Assist human proctors by highlighting moments for review

• Provide automated flags to streamline quality checks

These features improve proctor performance and enhance security, but they never replace human judgement.

2. What AI Does Not Do

To ensure fairness and compliance:

• AI does NOT make decisions about test integrity

• AI does NOT determine outcomes (e.g., pass/fail)

• AI does NOT issue sanctions or take action against candidates

• AI does NOT replace human proctors

• AI does NOT profile or score candidates

All AI outputs are advisory only.

Every flagged event is reviewed by a trained human proctor or exam administrator before any action is taken.

3. AI Is Not Used in:

• SmarTest Examiner (exam setup, marking, item authoring)

• SmarTest Nexus

• Test-centre operations

• Any identity verification workflows

• Any scoring or assessment decisions

CAS does not use AI to evaluate performance or behaviour beyond anomaly detection in SmarTest Invigilate.

4. Human Oversight and Governance

AI outputs are always reviewed and validated by:

• Trained proctors

• Operations managers

• Quality assurance personnel

We ensure:

• No automated decisions are made

• AI is used only as a support tool

• Humans remain responsible for all final decisions

• AI behaviours are monitored for accuracy and fairness

This approach aligns with UK GDPR, EU GDPR, and emerging guidelines for AI governance.

5. Data Protection & Transparency

We apply strict technical and organisational controls to protect candidate data, including:

• Encryption in transit and at rest

• Access controls and audit logging

• Data minimisation

• Retention limits

• Vendor due diligence

Further details are available in our:

• Privacy Policy

• Candidate & Client Privacy Notice

• Cookie Notice

AI-generated flags are treated as personal data and handled accordingly.

6. Future AI Use

Any new AI features:

• Will not be deployed without human oversight

• Will undergo data protection assessment (DPIA)

• Will be reviewed for fairness, transparency and proportionality

• Will be announced on this page before implementation

CAS is committed to responsible, ethical and transparent AI use.

7. Contact

For questions about our use of AI, please contact: Privacy@iagassessment.com.

Request more information