Artificial Intelligence Use in Certification Policy
Controlled use of ai without compromising fairness, validity, confidentiality or accountability.
Purpose
controlled use of AI without compromising fairness, validity, confidentiality or accountability
Scope
This policy applies to UKG personnel, contractors, subject-matter experts, assessors, decision-makers, administrators and candidates to the extent relevant to their role.
Mandatory controls
- AI use cases must be documented and risk-classified.
- AI-generated or AI-assisted outcomes must be monitored and validated for their intended use.
- High-impact assessment judgements require appropriate human oversight.
- No AI system may make an uncontrolled final certification decision.
- Candidate data must not be sent to an AI service without an approved lawful and security basis.
- Models, prompts, rules and material changes must be controlled and tested.
- Candidates should be informed where AI materially affects assessment or review.
Required records
- AI tool/version, purpose, validation record, human review and incidents.
Review and control
This controlled document is reviewed at least annually and whenever there is a material legal, scheme, technology, risk or accreditation change. The English master controls; translations must be reviewed after a material English change.