Independent evidence for accountable AI

MethodologyAboutAPI
AI GovernanceWorld
Primary navigation
MethodologyAboutAPI

Advanced course

Responsible AI Auditor

Advanced training for planning and performing evidence-based audits of AI governance, systems, controls, outcomes, and responsible-AI claims.

Curriculum

From principle to operating evidence

Lessons use concise analysis and realistic cases. Enrolled learners can open the full material and record completion.

Module 1

Audit foundations

Establish credible authority, criteria, scope, independence, and professional judgment.

  1. Assurance and the audit mandate · 40 minutes · Article
  2. Criteria, materiality, and scope · 40 minutes · Article
  3. Case study: auditing a responsible label · 40 minutes · Case study

Module 2

Planning and evidence

Design a risk-based audit program and obtain sufficient, appropriate evidence.

  1. Risk-based audit program · 45 minutes · Article
  2. Evidence, sampling, and reproducibility · 45 minutes · Article
  3. Case study: the perfect dashboard · 40 minutes · Case study

Module 3

System and control testing

Evaluate technical behavior and governance controls as an integrated socio-technical system.

  1. Data, model, and outcome testing · 50 minutes · Article
  2. Governance control testing · 45 minutes · Article
  3. Case study: human in the loop · 40 minutes · Case study

Module 4

Findings, conclusions, and follow-up

Communicate evidence, risk, and assurance with precision and verify sustainable remediation.

  1. Findings and root-cause analysis · 45 minutes · Article
  2. Conclusions and reporting · 45 minutes · Article
  3. Case study: closed on promise · 40 minutes · Case study

Learning outcomes

What you will be able to do

  • Define audit scope, criteria, objectives, materiality, independence, and competence for AI engagements.
  • Build risk-based audit programs that trace deployed systems from data and models to decisions and outcomes.
  • Evaluate control design, implementation, and operating effectiveness using reliable evidence and reproducible tests.
  • Assess fairness, performance, robustness, transparency, human oversight, suppliers, monitoring, and incident controls.
  • Write calibrated findings, conclusions, and follow-up evidence without overstating assurance.