Governance of Automated Decisions: Audit, Oversight and Control for AI Decision Models Training Course

Transformations When Decisions Become Model-Driven

• The shift from rules and human judgement to scored options.
• The speed and scale of automated actions.
• Emerging failure modes and accountability gaps.

Decision Inventory and Risk Classification

• Cataloguing automated decisions.
• Assessing impact, reversibility, and affected parties.
• Establishing risk tiers and corresponding control expectations.
• Defining ownership for each decision class.

Accountability and Human Oversight

• Distinguishing between meaningful and nominal human review.
• Treating autonomy thresholds as governance decisions.
• Defining escalation routes and response times.
• Identifying the accountable owner and sign-off process.

Decision Records and Explainability

• Components of a defensible decision record.
• Retention of confidence levels, options, and context.
• Traceability of models and versions.
• Explaining decisions to customers or regulators.

Policy and Boundary Control

• Managing approved option sets and change control.
• Governance of thresholds and periodic reviews.
• Defining prohibited decisions and hard stops.
• Implementing policy as code alongside model output.

Regulatory Expectations

• Obligations under the EU AI Act for decision systems.
• Sector-specific rules in finance, insurance, and health.
• Data protection standards and automated decision rights.
• Requirements for documentation and conformity evidence.

Vendor, Licence, and Provenance Risk

• Comparing hosted decision APIs versus open weights.
• Licence terms for derivative and embedded models.
• Assessing supplier dependency and discontinuation risk.
• Ensuring weight integrity and version pinning.

Bias, Fairness, and Outcome Monitoring

• Analysing outcome disparity across groups.
• Identifying proxy variables and option framing issues.
• Utilising appeal and complaint signals.
• Conducting periodic fairness reviews.

Incident Response for Incorrect Decisions

• Detecting systematic decision errors.
• Containing incidents and suspending autonomy.
• Managing remediation and customer redress.
• Performing root cause analysis and updating controls.

Assurance Review of a Decision Layer

• Identifying evidence auditors should request.
• Testing accuracy and calibration claims.
• Sampling decisions and overrides.
• Reporting findings to the board.