MLARM study guide
MLARM · Work through your exam, one topic at a time.
The Machine Learning and Artificial Intelligence Risk Management (MLARM) credential is designed for professionals who need to identify, assess, and mitigate risks introduced by AI and ML systems. It is relevant for risk managers, model validators, compliance officers, data scientists, and auditors working in regulated industries such as finance, healthcare, and technology. The MLARM study guide has 8 sections and about 2,046 words.
01Foundations of AI/ML and Risk Management
This domain checks whether you can distinguish traditional model risk from AI/ML risk and apply a risk-based lens across the lifecycle. Expect questions on ML paradigms, generalization, validation, risk tiering, and the limits of accuracy as a safety measure.
304 words, about 2 min02AI/ML Governance, Frameworks, and Regulation
Governance, frameworks, and regulation questions test whether you can map obligations to roles, documents, and controls, not just recite principles. Governance assigns accountability: the board and senior management own risk appetite; an AI or model risk committee sets policy; three lines of defense separate execution, oversight, and audit.
299 words, about 1 min03Data Quality, Privacy, and Security
This domain tests data fitness, privacy, and security across collection, training, and operation. Questions often combine legal concepts with technical failure modes.
239 words, about 1 min04Model Development, Validation, and Documentation
Development, validation, and documentation questions test independent challenge, evidence quality, and whether documentation supports decisions after deployment. Validation is independent of development and covers conceptual soundness, data quality, methodology, implementation, outcomes, sensitivity, and benchmarking.
244 words, about 1 min05Model Performance, Monitoring, and Drift
Performance and monitoring questions test metric selection, drift diagnosis, and response. You need to know what changed and why it matters.
242 words, about 1 min06Explainability, Fairness, and Bias
Explainability, fairness, and bias questions test whether you can choose the right explanation for the audience and interpret fairness metrics without overclaiming. Global explanations describe model behavior overall; local explanations justify an individual decision.
257 words, about 1 min07Operational Risk, Third-Party, and Model Lifecycle
Operational, third-party, and lifecycle questions test resilience and control across outsourcing, change, and retirement. The lifecycle runs from ideation and inventory through development, validation, deployment, monitoring, change, and decommissioning.
233 words, about 1 min08AI Ethics, Responsible AI, and Emerging Topics
Ethics and emerging-topic questions test principled judgment under uncertainty, especially for generative and agentic AI. Responsible AI principles include beneficence, non-maleficence, fairness, transparency, accountability, privacy, and human oversight.
228 words, about 1 min