Practical writing for people who run AI governance programmes. We aim for articles you could hand to a colleague as a working method — and we say where the evidence is thin.
Launch articles
Why AI governance starts with an inventory
Every other governance control depends on knowing what AI you run. Why inventories fail, what a minimum viable record looks like, and how to get to coverage in weeks rather than quarters.
Topics: inventory, risk tiering, shadow AI — EI-01, EI-02, EI-07, EI-09.
Runtime guardrails for AI agents: a practical rollout model
Agents act, not just answer. How to design least-privilege tool access and human approval for high-impact actions, and how to move from Observe to Enforce without breaking production — with Contain ready for incidents.
Topics: agents, runtime policy, human approval — IS-11, IS-12, IS-13, IS-15.
Readiness is not compliance: how to read framework mappings
A mapping from a control to a clause is a starting point for analysis, not a conclusion. How to use crosswalks across NIST AI RMF, ISO/IEC 42001 and the EU AI Act without overclaiming.
Topics: framework mappings, evidence labels, disclaimers — SE-05, AO-07.
Get new articles by email
Subscribe to the newsletter. One or two emails a month; unsubscribe at any time.