A retail bank is preparing to deploy a new AI tool that ranks loan applicants by predicted creditworthiness.
The model was trained on five years of historical lending data drawn from a single metropolitan region.
Before go-live, the AIMS lead must propose a fairness check consistent with ISO/IEC 42001 expectations on responsible AI. Which approach BEST reflects the standard ' s intent for bias detection during AI system impact assessment?

• A. Evaluate predicted-outcome disparities across protected groups using pre-defined fairness metrics, document the results in the AI System Impact Assessment, and confirm acceptance criteria are met before approving the model for deployment.
• B. Document an assumption in the AI System Impact Assessment that historical lending data is representative of the wider applicant population and proceed to deployment, with no quantitative fairness testing of model outputs prior to release.
• C. Restrict the deployment to a single business unit for the first month, with no formal pre-launch fairness testing, and rely on customer complaints to surface unfair outcomes for downstream investigation by the responsible model team.
• D. Wait for the first quarter of live decisions to accumulate, then run a retrospective fairness analysis on actual decision outcomes, since live data is more representative than any test data the bank could realistically assemble in a development environment.

A logistics company has implemented an AI system that recommends staffing levels for warehouse shifts. The user interface shows each recommendation with an ' Approve ' button. In practice, shift managers approve every recommendation within seconds because the system rarely surfaces context, and the interface does not let managers see the inputs that drove the decision. Reviewing this for responsible-AI conformance, which observation BEST reflects the human-oversight principle from ISO/IEC 42001 Annex B?

• A. The oversight is not meaningful because managers lack the information and context needed to challenge or override the recommendation, so the design provides the appearance of human control without the substance Annex B ' s human-oversight principle expects.
• B. Human oversight should be removed in favour of fully automated approvals on grounds of consistency, with sample audits performed monthly by the operations team to verify that the system ' s recommendations remain reasonable across the warehouse network over time.
• C. Human oversight is not engaged here because shift staffing is a low-risk operational decision; Annex B
  ' s oversight principle applies primarily to high-risk AI systems such as those that affect safety or fundamental rights, so the interface design issue is immaterial.
• D. The approve-button workflow satisfies human-oversight expectations because a human is in the loop for every staffing decision, and the speed of approvals reflects the efficiency gains of well-designed AI tooling rather than a control weakness in the deployed interface.

A documentation lead is preparing a one-page reference for new joiners that names the four annexes of ISO
/IEC 42001 and their purpose. She wants the description that aligns with GAICC Foundation Module 3.
Which option BEST matches the role of each annex in the standard?

• A. Annex A is a list of accredited certification bodies; Annex B is a Stage 2 audit checklist; Annex C is a Statement of Applicability template; Annex D is a financial impact assessment model.
• B. Annex A is a published list of certified consultants; Annex B is the official EU AI Act mapping; Annex C is a sample AI policy template; Annex D is the standard ' s revision history log.
• C. Annex A maps ISO/IEC 42001 to ISO 9001 directly; Annex B is a vocabulary appendix; Annex C is a sample Statement of Applicability; Annex D is a glossary of regional AI regulations.
• D. Annex A is the reference control set; Annex B is implementation guidance; Annex C catalogues AI- related objectives and risk sources; Annex D shows integration with other management system standards.

An AIMS project team is drafting the interested-parties register required under Clause 4.2. A team member suggests limiting the register to paying customers and regulators. The compliance lead pushes back, citing the breadth of interested parties expected under GAICC Foundation Module 4. Which option BEST captures the range of interested parties Clause 4.2 expects?

• A. Customers and competent regulatory authorities, since interested parties under Clause 4.2 are deliberately restricted to entities with formal commercial or statutory relationships.
• B. Top management and members of the AI governance committee, since interested parties under Clause
  4.2 refers to the internal leadership group accountable for the AIMS.
• C. External certification body auditors and the accreditation body, since interested parties under Clause 4.2 captures the assurance ecosystem that audits the AIMS.
• D. A broad range that includes customers, regulators, employees, data subjects, suppliers, the public, and communities affected by the organisation ' s AI activities.