WECH-RB Registered AI-GSR Specialist
AI Governance, Security & Responsible Use
Govern, secure and responsibly deploy AI systems, tools and agents across the enterprise.
Course curriculum
12 modules · 96-question timed exam (150 min, pass 72/96 = 75%) · human-reviewed capstone · target ~60h
1AI systems, model limits, RAG and agent lifecyclesPublished
Outcome: Explain and diagram AI/LLM/RAG/agent components and their failure modes for a given use case.
- Anatomy of an LLM feature: model, prompt, context, output — and their limits
- Diagnosing a retrieval-augmented assistant with evidence
- LLM-based agent lifecycles: tools, autonomy and where control must sit
3 lessons · 6 practice · 15 exam items · workshop included
2Governance charters, accountability, policy and risk appetitePublished
Outcome: Draft a governance charter mapping accountable roles to decisions and a stated AI risk appetite.
- From principles to a charter: accountable roles and decision rights
- Writing a usable AI policy: scope, prohibited uses and decision gates
- Risk appetite and tolerances: stating how much AI risk you will accept
3 lessons · 6 practice · 15 exam items · workshop included
3AI inventory, use-case scoping and stakeholder mappingPublished
Outcome: Build an AI inventory entry and scope a use case with stakeholders and intended/limited uses.
- The AI inventory: what to record, and why it is the backbone of oversight
- Scoping a use case: intended purpose, context of use and limited uses
- Stakeholder mapping and characterizing impacts
3 lessons · 6 practice · 15 exam items · workshop included
4Impact assessment, risk prioritization and human-rights impactsPublished
Outcome: Conduct an AI impact assessment and prioritize risks including human-rights impacts.
- The AI impact assessment: purpose, components and when it is required
- Prioritizing risks by severity, likelihood and who is affected
- Human-rights and disparate-impact harms
3 lessons · 6 practice · 15 exam items · workshop included
5Data governance, privacy, provenance and retentionPublished
Outcome: Design a data map with provenance, privacy basis and retention for an AI use case.
- Data mapping and provenance for AI
- Privacy basis, data minimization and sensitive data
- Retention, deletion and the data lifecycle
3 lessons · 6 practice · 15 exam items · workshop included
6Evaluation design, reliability, bias and human oversightPublished
Outcome: Design an evaluation plan covering reliability, bias and effective human oversight.
- Evaluation Design for AI System Reliability
- Identifying and Mitigating Bias in AI Systems
- Effective Human Oversight and AI Governance
3 lessons · 6 practice · 15 exam items · workshop included
7LLM threat modeling, prompt injection and output handlingPublished
Outcome: Threat-model an LLM feature and specify prompt-injection and output-handling controls.
- Threat Modeling an LLM Feature as a Sociotechnical System
- Prompt Injection Defense and Tool-Use Constraints
- Output Handling, Disclosure Controls, and Operational Monitoring
3 lessons · 6 practice · 15 exam items · workshop included
8Agent/tool security, least privilege, approval and containmentPublished
Outcome: Specify least-privilege, approval and containment controls for an agent with tools.
- Tool Authorization Boundaries and Least Privilege
- Approval Architecture for Consequential Agent Actions
- Containment, Monitoring, and Incident Response for Agentic Systems
3 lessons · 6 practice · 15 exam items · workshop included
9Third-party/model/vendor due diligence and contractsPublished
Outcome: Assess an AI vendor/model and define contract and assurance requirements.
- Risk-Based AI Vendor and Model Due Diligence
- Contract Controls, Assurance Rights, and Operational Accountability
- Continuous Assurance, Monitoring, and Vendor Change Governance
3 lessons · 6 practice · 15 exam items · workshop included
10Deployment gates, validation, change and release controlPublished
Outcome: Define deployment gates and change/release controls with validation evidence.
- Deployment Gates: Converting Risk Decisions into Release Criteria
- Validation Evidence for AI Releases
- Change Control, Release Integrity, and Post-Deployment Revalidation
3 lessons · 6 practice · 15 exam items · workshop included
11Monitoring, incident response, drift and recoveryPublished
Outcome: Design monitoring, drift detection and an AI incident-response and recovery playbook.
- Monitoring architecture, performance signals, and drift triage
- AI incident response: detection, containment, investigation, and communications
- Recovery, revalidation, governance reporting, and resilience
3 lessons · 6 practice · 15 exam items · workshop included
12Assurance, audit evidence, metrics and governance improvementPublished
Outcome: Assemble audit evidence and governance metrics and plan improvement actions.
- Building Defensible AI Audit Evidence
- Metrics That Inform AI Governance Decisions
- Governance Findings, Remediation, and Continuous Improvement
3 lessons · 6 practice · 15 exam items · workshop included
Curriculum v2.2026.11 — published by WECH-RB and authored in-house; not independently accredited. Enrolled learners get the full course now. Earning the credential requires passing the timed exam and the human-reviewed capstone and holding a current annual membership.
Fees & exam rules
Credential awarded: WECH-RB Registered AI-GSR Specialist
- Pass 72/96 (75%) — multiple-choice, auto-graded
- 150-minute timed online exam, 96 questions (8 per module), safe resume
- One annual membership ($125/yr) keeps every credential you earn valid
- Plus a human-reviewed capstone (pass 75/100)
Before you enroll
These materials are published by WECH-RB and authored in-house; they are not independently accredited. The certification exam for this track is still in development — enrolling provides course access only and does not award a credential.
$595
Course fee, one-time. Exam purchased separately.
You'll sign in to enroll.