Manual review and exception handling

Manual review is not a magic fallback. It needs defined evidence, trained reviewers, consistent outcomes, privacy controls, escalation and a route for the applicant to correct mistakes or use an alternative.

For: identity operations, fraud, privacy, customer support and procurement teams. This is independent information, not legal, compliance or security advice.

Design the queue

• Define which automated outcomes enter review.
• Prioritise by risk and user harm, not only arrival time.
• Show reviewers the minimum evidence needed.
• Separate document, biometric, fraud and eligibility judgements.
• Prevent one reviewer from silently changing policy.

Reviewer controls

• Vetting, training and recurrent assessment.
• Role-based access and monitored evidence viewing.
• Decision reasons and controlled free text.
• Escalation for uncertainty, suspected fraud and vulnerable users.
• Quality sampling, disagreement review and feedback.

Applicant experience

Explain that review is pending, give a realistic timeframe, avoid repeated unnecessary uploads, and provide correction and appeal. Design for name variations, changed appearance, disability, worn documents and people without the default evidence.

Measure the real operation

Track queue time, reversal, reviewer agreement, fraud detection, complaints, data access and outcomes by relevant segment. A low manual-review rate can hide aggressive automated rejection rather than better automation.

Evidence and limits

MyID separates enacted rules, official implementation material, testing and vendor claims. A source can establish what its publisher says; it does not prove that every product, deployment or interpretation works as claimed. Where Irish implementation remains unsettled, this page says so.

• NIST SP 800-63A-4 identity proofing
• General Data Protection Regulation

Sources checked 22 August 2026. Re-check the linked primary material before making a consequential decision.

Next useful pages

• Accuracy and fairness
• Document verification
• DPIA and redress questions