AEGIS · Module 5 · Cadence
Artifact 5.3
Role-Based Training Curriculum
Five tiers of required training matched to how each role actually touches AI — with competency checks, renewal cadence, and audit-grade completion records.
- Client
- [CLIENT NAME]
- Engagement
- [ENGAGEMENT ID]
- Version
- v1.0
- Issued
- 2026-05-18
Delivered by TechFides under the AEGIS Governance Operating Services engagement. This document is proprietary to the client named above. Redistribution beyond the engagement steering committee requires written consent.
Purpose
This curriculum defines the minimum AI literacy and competency required for every role in the organization, matched to the level of exposure that role has to AI systems and AI-assisted work. It is the evidentiary backbone for EU AI Act Article 4 (AI literacy), NIST AI RMF GOVERN-1.6 (workforce competencies), ISO 42001 Clause 7.2, and the HR dimension of ISO 27001 A.6.3. Completion records are retained for audit and refreshed annually.
Five Training Tiers
Intent — Tiers are additive. A Builder completes Baseline + Practitioner + Builder. A Governance role completes Baseline + Practitioner + Governance. Executives complete Baseline + Executive.
Audience: Every employee, contractor, and long-term vendor staff with systems access.
Learner will be able to
- Can explain in one sentence what the Acceptable Use Policy permits and prohibits.
- Knows the list of approved AI tools and where to check the current inventory.
- Knows what data classes may never be pasted into a non-approved AI tool.
- Knows how to report an AI-related incident or concern.
Audience: Any employee whose daily workflow includes an AI-assisted task (per Artifact 4.1).
Learner will be able to
- Can identify which autonomy tier (T1–T4) a workflow operates under and what their review obligation is.
- Can produce a high-quality prompt for a workflow using the approved template (Artifact 4.2).
- Can recognize the top 3 failure modes for their workflow and knows when to escalate.
- Can explain the human-in-the-loop role for the workflow they run most often.
Audience: Anyone who authors prompts, designs workflows, integrates AI into systems, or configures agents.
Learner will be able to
- Can design a new workflow using the 11-field spec (Artifact 4.1) and route it through the approval path.
- Can write an evaluation set + red-team prompt suite for a new workflow (Artifact 4.2).
- Can map a new workflow to relevant frameworks and data classifications.
- Can articulate the choice of autonomy tier and defend it at QGR.
Audience: AI Governance Lead, CISO, General Counsel, Compliance, Privacy, Internal Audit, Data Protection Officer.
Learner will be able to
- Can run the full Risk Register review (Artifact 2.3) including scoring recalibration.
- Can chair an AI incident (Artifact 2.6) from detection to post-mortem.
- Can map a new workflow to NIST AI RMF, ISO 42001, EU AI Act, and surface gaps.
- Can defend the program to an external auditor with evidence pulled in under 24 hours.
Audience: Executive Sponsor, CEO, CFO, COO, Board AI Committee members.
Learner will be able to
- Can explain the governance posture of the organization in one paragraph.
- Can interrogate an AI business case against the Value & Spend Tracker (Artifact 3.3).
- Can make Scale / Hold / Rework / Retire calls at QGR with confidence.
- Can represent the program credibly to customers, regulators, and the board.
Role → Tier Assignment
Intent — Every role in the org chart maps to one or more tiers. HR and the AI Governance Lead maintain this mapping jointly and review it quarterly.
| Role / Function | Required Tiers | Trigger |
|---|---|---|
| All staff + contractors | Baseline | On or before Day 1. |
| Sales, CS, Support, Ops, Finance, Marketing staff using AI | Baseline + Practitioner | Before first access to an AI-assisted workflow. |
| Engineering, Data, ML, Product using AI assistants or agents | Baseline + Practitioner + Builder | Before authoring any prompt, workflow, or integration. |
| AI Governance Lead, CISO, GC, Compliance, Privacy, DPO, Internal Audit | Baseline + Practitioner + Governance | Before taking the role; refresher quarterly. |
| Executive Sponsor, CEO, CFO, COO, Board AI Committee | Baseline + Executive | Before taking the role; refresher annually. |
| Third-party / vendor staff with tool access | Baseline (attested by vendor) | Before provisioning access. Re-attest annually. |
Competency Checks
Intent — Each tier ends with a check. Pass threshold is 80% unless noted. Failure pathway is the same across tiers: retake once, then instructor-led remediation before a second attempt.
Baseline
Method: 10-question multiple-choice, randomized
Pass rule: 8/10
Evidence: LMS completion record + score + timestamp. Retained 3 years.
Practitioner
Method: Scenario: classify 5 real (redacted) prompts against data-class + autonomy-tier rules. Short-answer: escalation path for named workflow.
Pass rule: 4/5 scenario + 3/3 short-answer
Evidence: LMS record + graded rubric + instructor sign-off.
Builder
Method: Portfolio piece: submit a full workflow spec (Artifact 4.1) + prompt (Artifact 4.2) for a fictional case. Reviewed by AI Governance Lead against rubric.
Pass rule: Meets 10/12 rubric criteria; 2 attempts max before 30-day cool-off.
Evidence: Portfolio retained + reviewer signature + ADR-style rationale.
Governance
Method: Tabletop: chair a simulated incident + Risk Register recalibration. Observed by external coach.
Pass rule: Pass/fail by coach rubric (8 criteria). Fail = 1:1 coaching + re-run within 60 days.
Evidence: Coach report + tabletop artifacts + HR file.
Executive
Method: Decision simulation: Scale / Hold / Rework / Retire call on 3 scripted cases with trap-doors (unclear metrics, conflicting value cases).
Pass rule: Defensible rationale on all 3 per observer.
Evidence: Observer notes + HR file + re-run at annual refresh.
Renewal & Currency
Intent — A trained workforce a year ago is not a trained workforce today. These cadences are the minimum.
| Tier | Full Renewal | Mandatory Re-Run Triggers | Responsible |
|---|---|---|---|
| Baseline | Annual | New AUP version · after any material incident | HR + AI Governance Lead |
| Practitioner | Annual + on-workflow-change | New or materially changed workflow · autonomy-tier change | Function Head + AI Governance Lead |
| Builder | Annual + on-model-change | New foundation model · new data-class in scope · new integration pattern | AI Governance Lead |
| Governance | Quarterly 60-min refresh + annual full | New regulation · new framework version · cross-border expansion | AI Governance Lead + GC |
| Executive | Annual + pre-QGR briefing | New board member · material strategy change · material incident | AI Governance Lead + Exec Sponsor |
Module Inventory
Intent — The modules that compose this curriculum. Numbered, owned, versioned. Stored in LMS with source-of-truth in the governance repo.
| Module | Tier | Owner | Version |
|---|---|---|---|
| M-01 · AI at [Client]: What, Why, Guardrails | Baseline | AI Governance Lead | [v1.0] |
| M-02 · Acceptable Use in Practice | Baseline | AI Governance Lead + GC | [v1.0] |
| M-03 · Data Classes + What You May Never Paste | Baseline | CISO + DPO | [v1.0] |
| M-04 · Reporting a Concern | Baseline | AI Governance Lead | [v1.0] |
| M-05 · Prompting for Quality | Practitioner | AI Governance Lead | [v1.0] |
| M-06 · Autonomy Tiers + Human-in-the-Loop | Practitioner | AI Governance Lead | [v1.0] |
| M-07 · Escalation + Failure Modes | Practitioner | Function Heads | [v1.0] |
| M-08 · Workflow Design Workshop | Builder | AI Governance Lead | [v1.0] |
| M-09 · Evaluation + Red-Teaming | Builder | AI Governance Lead + ML Lead | [v1.0] |
| M-10 · Framework Mapping Seminar | Governance | AI Governance Lead + Compliance | [v1.0] |
| M-11 · Incident Tabletop | Governance | CISO + AI Governance Lead | [v1.0] |
| M-12 · Executive Decision Simulation | Executive | Executive Sponsor + AI Governance Lead | [v1.0] |
Evidence & Audit
Intent — What a regulator or auditor will ask for, and where we hold it.
Show that every employee completed AI literacy training this calendar year.
Evidence: LMS completion report, exportable per person and per cohort. Matches HRIS active-employee list with ≤2% drift.
Owner: HR + AI Governance Lead
Show that every Practitioner passed the competency check before touching a workflow.
Evidence: Workflow access provisioning log tied to LMS pass record. Access is denied in IAM until LMS status = pass.
Owner: IT / IAM + AI Governance Lead
Show that Builders cannot author a production prompt without Builder certification.
Evidence: Repo merge protection + code-owner rule. Only users in the AI-Builders group can merge to prompt library.
Owner: Engineering + AI Governance Lead
Show that Governance roles practice incidents before a real one happens.
Evidence: Tabletop schedule + after-action reports for the last 4 quarters.
Owner: CISO + AI Governance Lead
Show that Executives made decisions at QGR with up-to-date training.
Evidence: Annual Executive refresh sign-off matched to QGR attendance log (Artifact 5.1).
Owner: AI Governance Lead + Exec Sponsor