What is the GARP RAI exam?
The Risk and AI certificate is GARP's newest credential, announced in June 2024 with its first exam window that November. It covers how machine learning and AI tools work, where they are being applied in risk management, the new risk factors they introduce, what responsible and ethical AI means in a financial context, and how firms govern data and AI models.
Like the SCR it is a certificate rather than a designation, with a single exam, no work-experience requirement and no prerequisites. GARP designed it to be approachable for candidates from mixed backgrounds rather than only for quantitative specialists, and frames its questions in realistic workplace scenarios.
Its syllabus has six modules, of which the last, a set of case studies and practitioner perspectives, is optional and not examined.
Who sits it
- Risk professionals whose firms are deploying machine learning in credit, market or operational risk
- Model risk and model validation teams facing AI models for the first time
- Compliance, audit and governance staff who must oversee AI systems
- FRM holders adding an AI credential, who get a reduced exam fee
Exam format at a glance
- Type
- A certificate, not a designation. GARP's only designation is the FRM
- Questions
- 80 equally weighted multiple-choice questions; mostly standalone, with some groups sharing a scenario
- Duration
- 4 hours
- Delivery
- Pearson VUE, either at a test centre or online proctored
- Exam windows
- Twice a year, in April and October
- Fees (USD)
- Around $550–650 early and $650–750 standard depending on membership; FRM and SCR holders pay the lowest tier. Retakes are $350
- Prerequisites
- None. No work experience is required
- Passing standard
- Reported as pass or fail with quartile feedback; GARP does not publish a numeric passing score
- Recommended study
- GARP estimates 100–130 hours
- First sitting
- November 2024
Exam-body facts last checked September 2026. Fees, dates and formats change; confirm the current details with Global Association of Risk Professionals (GARP) before registering.
Syllabus and topic weights
GARP publishes the number of questions drawn from each chapter inside its downloadable study guide rather than percentage weights on its public pages, so no percentages are shown here. Note that the sixth module is optional and carries no questions.
- 1.AI and Risk: Introduction and Overview
- 2.Tools and TechniquesThe largest module, covering the main families of machine learning method
- 3.Risks and Risk Factors
- 4.Responsible and Ethical AI
- 5.Data and AI Model Governance
- 6.Case Studies and Practitioner PerspectivesOptional and not examined
How to prepare for GARP RAI
- 1
Do not skip the governance and ethics modules in favour of the technical one. Two of the five examined modules are about responsible AI and model governance, and they are where candidates from engineering backgrounds tend to lose marks.
- 2
Tools and Techniques is the largest module and the one to start with, since the risk-factor material assumes you know how the methods work.
- 3
Ignore the sixth module for exam purposes. It is case studies and practitioner perspectives, and GARP states it is not tested.
- 4
Expect practical framing. Questions are set in realistic work scenarios rather than as abstract definitions, so passive reading is poor preparation.
- 5
This is a young exam, so there is less third-party practice material than for established credentials. GARP's own curriculum and practice exams are included in the fee.
We don't cover GARP RAI yet
We have no Risk and AI study material and none planned. Our FRM Part 2 content covers model risk, validation and the quantitative methods this certificate builds on, and GARP gives FRM holders a reduced fee for it.
GARP RAI FAQ
How new is this certificate?
GARP announced it in June 2024 and ran the first exam that November, so there is only a short history of sittings and relatively little third-party study material.
Do I need a technical background?
No. GARP states it is designed to be accessible to candidates from a wide range of backgrounds, and it is not a machine-learning engineering exam. It does assume you are willing to understand how the methods work.
Is it worth taking alongside the FRM?
They serve different purposes. The FRM is the substantive risk designation; this is a focused certificate on one fast-moving topic. FRM holders do pay the lowest exam fee tier for it.
Does StudyOptima have Risk and AI questions?
No, and none are planned at present. Our FRM Part 2 material covers model risk and validation, which is the closest overlap.