Layer 2 · Measurement ecosystem · Adjacent · Observation · Organization

OLD Infrastructure and Culture Audit (OLD-ICA)

An auditor-led review of documents, practices and interviews that explains why an organization carries the learning debt the OLD-MA identifies.

Status
Draft for development (v1.0)
Version
v1.0 draft (February 2026)
For
Trained auditors, organizational leads for AI literacy and AI governance
Page updated
October 2026

What it is for

The OLD Maturity Assessment tells an organization how much learning debt it has accumulated. The OLD Infrastructure and Culture Audit tells it why: which infrastructure gaps, cultural conditions and governance deficiencies are producing and sustaining that debt. The OLD-MA gives the dimension profile; the audit gives the explanatory narrative. Neither is complete without the other.

Classification

Adjacent / Observation / Organization. It is Adjacent because it measures the conditions that produce learning debt, not the debt itself or ALiF competencies. It uses Observation because a trained auditor reviews documents and practices against a structured protocol. Its unit of analysis is the organization.

What it audits

The audit has four domains that mirror the four debt dimensions:

Domain Focus
Capability infrastructure How the organization defines, assesses, tracks and updates AI literacy by role, and whether competency planning accompanies new AI deployments
Learning infrastructure Feedback and incident reporting on AI use, evaluation of training and AI systems, communities of practice, protected learning time and budget
Culture and leadership How leaders talk about AI, recognition of AI learning, psychological safety for raising AI concerns, and leadership participation in learning
Governance and pipeline Whether AI governance is active and connected to practice, frontline representation, and deliberate design of early-career pathways where AI has taken over work that used to build expertise

Each domain combines a document review (for example training records, governance minutes, incident reports and learning-time schedules) with key informant interviews held separately with people at different levels. A requested document that cannot be provided counts as evidence of absence.

How it works

The OLD-MA workshop comes first. The auditor then reviews its gap profile, the organization’s AI portfolio and the requested documents, and focuses deepest scrutiny on the dimensions with the largest gaps while still covering all four. Self-audit is not permitted: the auditor is external, or internal but independent of the workshop and of line management for the areas assessed, and has completed auditor training.

The audit ends with an integrated synthesis: patterns that cut across domains, a check of whether the evidence supports each OLD-MA rating or suggests it is too high or too low, and a narrative on what changes would most reduce debt accumulation.

Ethics

The audit requires institutional governance approval and informed consent from every person interviewed. Interview notes are stored separately and are not attributed to named individuals without consent.

Status

Version 1.0 is a draft for development, built in parallel with the OLD-MA for paired use. If the debt dimensions change after peer review of the concept paper, the audit domains will change in the same version cycle. A module within the audit focused on pipeline debt and learning conditions is planned. The protocol, checklists and interview guides are not published on this site.

Documents

Documents will be available to download when they are released under an open licence. Until then, contact the ALiF core team.