Skip to main content

Gap Closure Program

The AEEF framework provides standards, controls, and operating guidance. Most organizations still fail in implementation because they stop at documentation and do not convert standards into a tracked remediation program with owners, deadlines, and evidence.

This page closes that gap.

Use it with the repository root TODO.md, which is a framework-generic remediation backlog template.

For sovereign and multi-country positioning work (annex governance, coverage matrix, conformance terminology, and claim-boundary remediation), use the companion remediation plan at TODO-middle-east-sovereign-standards-remediation.md.

When to Use This Page

Use this page when you need to:

  • convert a review or audit finding into a structured action plan
  • implement AEEF controls in a real organization with evidence
  • prioritize foundational controls before optimization work
  • apply sector or regional overlays (for example banking, KSA, EU) without rewriting the core program

What This Adds to the Framework

This implementation layer provides:

  • a remediation program model (P0/P1/P2)
  • a core vs. overlay approach for reuse
  • execution templates for control catalogs, audit packs, CAPA, and simulations
  • a remediation tracker schema for consistent implementation evidence

Core vs. Overlay Model

Do not run every workstream at full depth on day one. Split implementation into:

Core Program (Always Applicable)

These actions are foundational in almost all environments:

  • control catalog and ownership
  • evidence model and retention linkage
  • internal audit cadence
  • release gates and runtime reliability controls
  • incident response and tabletop readiness
  • architecture baselines and vendor concentration review
  • organizational roles, RACI, and approval SLAs

Overlay Programs (Context-Driven)

Enable overlays only when applicable:

  • Regional/Jurisdiction overlays: KSA, EU, UAE, etc.
  • Sector overlays: banking/finance, healthcare, public sector
  • Channel/platform overlays: messaging, voice, app store ecosystems, marketplaces
  • Deployment overlays: sovereign/in-country, multi-tenant, regulated cloud

For multi-country or sovereign deployments, validate profile availability and gaps before activating overlays:

Step 1: Start the Program (Week 1)

  1. Appoint a remediation sponsor.
  2. Assign workstream leads.
  3. Create the remediation tracker.
  4. Define the implementation "definition of done" (policy + technical control + monitoring + evidence + test).

Step 2: Build the Core Backlog (Days 1-2)

  1. Use TODO.md to select P0 items.
  2. Translate each item into tasks with owners and dates.
  3. Attach evidence fields up front (not later).

Step 3: Choose Overlays (Days 1-2)

  1. Complete the overlay selection template.
  2. Activate only the relevant overlay tasks.
  3. Add overlay-specific evidence requirements to the tracker.
  4. If required annexes are missing, log the gap explicitly and create an annex/remediation work item rather than assuming coverage.

Step 4: Implement and Validate (Days 3-10)

  1. Implement technical and process controls.
  2. Automate evidence collection for high-frequency controls.
  3. Run simulations (audit, incident, breach, hallucination/high-impact failure).
  4. Convert findings into CAPA and track to closure.

Required Program Artifacts (Templates)

Use the templates under templates/program/:

  • templates/program/ai-control-catalog-template.md
  • templates/program/ai-use-case-inventory-template.md
  • templates/program/jurisdiction-applicability-register-template.md
  • templates/program/ai-audit-evidence-pack-template.md
  • templates/program/ai-nc-capa-register-template.md
  • templates/program/ai-tabletop-report-template.md
  • templates/program/ai-reference-architecture-blueprint-template.md
  • templates/program/board-ai-review-pack-template.md
  • templates/program/profile-overlay-selection-template.md

The remediation tracker payload schema is in:

  • templates/schemas/ai-remediation-program.schema.json

Recommended additions for sovereign/multi-country programs:

  • templates/program/regulator-objection-log-template.md (when available)
  • templates/program/national-annex-control-matrix-template.md (when available)

How This Maps to Existing AEEF Components

Use this page as an implementation wrapper around the existing framework:

Program NeedPrimary AEEF Components
Governance controls and evidencePillar 2: Governance & Risk
Regional claim boundaries, annex governance, coverage scopeRegional Scheme Governance, National Annex Specification, Regional Coverage Matrix
Runtime trust/reliability controlsPRD-STD-010, PRD-STD-012
Model/data/privacy governancePRD-STD-011, PRD-STD-014
Multi-tenant and channel overlaysPRD-STD-013, PRD-STD-016
Organizational readinessPillar 5: Organizational Enablement
AI product lifecycle implementationAI Product Lifecycle
Transformation rollout sequencingTransformation Implementation Hub

90-Day Minimum Outcome (What "Implemented" Looks Like)

By day 90, a serious implementation SHOULD be able to show:

  • a live AI use-case inventory and risk classification
  • a control catalog with named owners and evidence sources
  • at least one complete audit evidence pack for a high-priority AI system
  • runtime SLOs, fallback, and on-call ownership for in-scope production AI features
  • one mock audit and one incident tabletop with CAPA tracked to closure

Common Failure Modes (Avoid These)

  • Policy-only completion: marking items done when only a document exists
  • No evidence linkage: controls implemented but not provable
  • Overlay overload: enabling every jurisdiction/sector overlay before core controls are stable
  • Manual bottlenecks: too many approvals without automation or SLAs
  • No validation: skipping simulations and discovering failures during real incidents

Next Steps

  1. Open the repository root TODO.md and select the P0 core actions for your next 90 days.
  2. Complete templates/program/profile-overlay-selection-template.md to activate only applicable overlays.
  3. Create your control catalog and remediation tracker before implementing new controls.
  4. Schedule a mock audit and one incident tabletop before declaring the rollout "ready."