← Back to portfolio
Enterprise Lab 01 · Governance Infrastructure

Enterprise AI Governance & Risk Management Wiki

The policy and control repository for AI model risk in a financial-services organization, the governance backbone that defines how every model is validated, monitored, escalated, and reviewed across its lifecycle.

SR 11-7 · Model Risk Basel · Stress Testing EU AI Act · Fairness & Risk Tiering

01What this lab is

Governance is not documentation written after the fact.

It is the standing control layer that decides what is validated, who is accountable, when humans review, and how evidence is retained, before a model ever reaches production. This wiki is that layer for an organization expanding AI into a stronger financial presence: the single, transparent record of model practices, regulation guidelines, fraud-detection guidelines, audit history, and the risk-management procedure.

A risk that is named and mapped is a risk that is governed. An unnamed risk is the one that reaches production unchecked.

02What it governs

Section 01
Risk Assumptions
The known failure scenarios, seasonal fraud, bias in loan approval, data drift, the framework is built to catch.
Section 02
Validation Methods
Back-testing, benchmarking against challenger models, fairness analysis, and Basel stress testing on a 6-12 week mark.
Section 03
Data Lineage
A transparent record of data sources, transformations, quality checks, and the encryption and privacy standards applied.
Section 04
Escalation Rules
Who escalates, on what triggers (6% risk-flag breach, 0.7% drift), and how, within a 48-hour window.
Section 05
Review Cadence
Review frequency scaled to risk tier, the core of a mature model-risk function.
Section 06
Regulatory References
Every control traced back to SR 11-7, Basel, or the EU AI Act, with retained audit artifacts.

03Review cadence, scaled to risk

The defining principle: the higher the risk a model carries, the more often it is reviewed.

High Risk
Every 3 months
Loan risk model & credit applications
Medium Risk
Every 6 months
Credit-line inquiry systems
Standard
Once a year
Organization-wide SOPs

04How it works with Lab 02

Policy defines, tools execute

This wiki is the policy and control repository, it defines what policies apply, who owns decisions, how AI systems are classified, and when human review is triggered. The AI Transformation Portfolio Office (Lab 02) executes those rules against live initiatives, flagging governance exceptions and enforcing the risk-tiered review cadence across the whole portfolio. Together they form the governance operating system.

Enterprise AI Governance & Risk Wiki
The full governance framework, six sections, aligned to SR 11-7, Basel, and the EU AI Act.
View the wiki →