Technical practice
How would you turn AI governance principles and regulatory expectations into usable standards and procedures?
What to coverThe role supports the bank’s AI governance framework, policies and practical guidance. Suggested preparation: explain how you would translate expectations into documented controls, assign responsibilities and review the guidance as expectations change.
What lifecycle controls would you prioritise in a Model Management Platform?
What to coverThe posting identifies lifecycle controls, documentation, workflow automation and monitoring as platform capabilities. Suggested preparation: discuss how those capabilities could create a traceable path from use-case registration through review, approval and ongoing oversight, while keeping workflows usable for development teams.
How would you maintain an AI inventory and facilitate risk, materiality and FEAT assessments?
What to coverThe role includes managing the department’s AI inventory and facilitating all three assessments. Suggested preparation: describe the information you would collect, how you would handle incomplete entries, and how assessment outcomes would inform review priorities.
How would you validate a conventional AI model in a regulated environment?
What to coverPractical model validation and testing are required. Suggested preparation: cover intended use, data and assumptions, performance testing, limitations, independent challenge and the evidence needed to support a risk decision.
How would you design an evaluation for a Generative AI use case?
What to coverApplying model-risk principles to Generative AI is required, while hands-on evaluation design is preferred. Suggested preparation: explain how you would select representative and adverse test cases, judge output quality and safety, record failures, and decide what residual risk remains.
What additional risks would you test in an agentic AI system?
What to coverThe posting explicitly includes agentic AI within emerging use cases requiring model-risk assessment. Suggested preparation: consider tool access, action boundaries, unexpected multi-step behaviour and human oversight, and explain how testing and controls would address each risk.
How would you assess bias, fairness and explainability for a consequential AI use case?
What to coverThe role requires understanding of bias, fairness and explainability. Suggested preparation: discuss relevant groups and outcomes, the limits of available data and explanations, and how findings would translate into controls or a challenge to the implementation.
What would you monitor after an AI system is deployed?
What to coverMonitoring appears in both the required AI-lifecycle knowledge and the proposed platform capabilities. Suggested preparation: distinguish performance, fairness and control-effectiveness signals, then explain how thresholds, investigation and escalation could support ongoing oversight.
How would you review and challenge a critical or high-risk AI use case?
What to coverThe role reviews these use cases and recommends controls where implementations need challenge. Suggested preparation: explain how you would examine purpose, system behaviour, testing evidence, guardrails and residual risk, then communicate a proportionate recommendation.
How would you prepare an AI governance update or response to a regulator?
What to coverThe posting includes governance reports, management updates, regulatory submissions and regulator responses; familiarity with MAS and related risk frameworks is preferred. Suggested preparation: describe how you would verify evidence, distinguish facts from unresolved issues, and present decisions and remediation clearly without assuming a particular regulatory template.
Behavioural practice
Tell me about a time you challenged a high-risk technical implementation.
What to coverUse a real situation in which you identified a material concern, then explain the evidence you used and how you raised it. Describe the resulting decision or control, including any disagreement you had to resolve.
Describe a time you improved adoption of a governance workflow or platform.
What to coverSet out the initial adoption problem, what you did with development-team feedback, and how you balanced usability with oversight. Give the observable result rather than assuming that rollout alone meant success.
Tell me about a time you explained a complex governance requirement to non-technical stakeholders.
What to coverIdentify the audience and the decision or action they needed to take. Explain how you adapted your training or advice and what changed as a result.
Describe a time you produced decision-ready documentation under scrutiny.
What to coverUse a real management or regulatory-facing example, if you have one, and state the situation, your contribution and the outcome. Focus on how you checked accuracy, handled uncertainty and made the key risks understandable.
Tell me about a time changing expectations led you to revise a risk or governance approach.
What to coverExplain what changed, how you assessed its relevance and whom you involved before acting. Describe the resulting change to guidance or controls and how you checked that it was understood.