AI governance for high-stakes decisions

Govern AI before it governs the outcome.

We help mid-sized, regulated and public-interest organizations identify, govern and prepare for the responsible adoption of practical AI systems.

GR-0001 · GOVERNANCE RECORD
Establish governance
Identify & assess risk
Design & implement controls
Generate & retain evidence
Assure & report
Evidence
Risk register
v1.2 · 14 items
Evidence
Control library
v0.7 · 9 mapped
Evidence
Audit trail
v2.1 · 4 reviewed
Methodology informed by
NIST AI RMFISO/IEC 42001EU AI ActOECD PrinciplesAU Continental AI Strategy
01  The governance gap

Adoption moved first. Governance is still catching up.

AI is already inside procurement, operations, client service and employee workflows. Most organizations cannot say which systems are running, who owns them, what data enters them, which decisions depend on them, or what evidence exists that any of it was reviewed.

That is not a technology problem. It is an accountability problem — and it is answerable.

02  Governance Incident Register

Real failures. Dated and sourced.

A maintained record of governance failures in AI systems — what happened, which control failed, and what it means for organizations running something similar. Every entry traces to a primary source: a lab, a regulator, a court, or the disclosing organization itself. Never an aggregator.

REG-2026-10-22Revision pending

Railway platform breach — encrypted secrets decrypted with the credentials that reached them

Control failed: secrets were encrypted at rest, and the credentials able to decrypt them were reachable from the same build path the attacker compromised. Around 100,000 workspaces may have been exposed — belonging to organizations that were never themselves attacked.

Could we list today, without asking anyone, every credential we have placed in a third-party platform?

E4 · Heightened→ STRATA
REG-2026-10-21Verified

ConnectWise supply chain compromise — standing third-party trust at agent speed

Control failed: access granted once to a third-party application was standing, unattended and broader than its purpose. When the vendor was compromised, that consent carried the attacker into customer tenants with no interactive sign-in — over 200 sign-ins in under four minutes on four application credentials.

Which third-party applications can act inside our systems without a person signing in?

E4 · Heightened→ ATLAS · STRATA
REG-2026-09-24Verified

DIVD helpdesk breach — two zero-days chained to root by an automated agent

Control failed: the interval between an unknown vulnerability and root access was measured in seconds, which excludes every control that depends on someone reacting inside it. What limited the damage was structural and already in place — segmentation, not detection.

Which of our controls would still work if the attack finished before anyone read the alert?

E4 · Heightened→ STRATA
03  Method

Policy to proof to practice.

Three connected practices. One maintained record. Governance that can be reviewed rather than asserted.

Atlas

Governance & policy

Ownership, decision rights, intake and approval, policy framework, accountability mapping.

What rules and responsibilities govern our AI?
Strata

Risk & control readiness

Risk registers, control mapping, evidence requirements, vendor review, incident readiness.

What could go wrong — and what proves our safeguards work?
Nexus

Literacy & readiness

Role-based guidance, escalation paths, knowledge checks, documented preparation.

Do our people understand and apply those rules?
The Governance Record

One record that stays current — not three reports that age.

Exposure, readiness and evidence are reported separately and never blended into a single score. A single score would be easier to sell, and would not be true.

SystemsInventory, owners, exposure classification
PolicyRules, decision rights, approvals
ControlsRisks, safeguards, exceptions
EvidenceDeclared · referenced · reviewed
PracticeRoles, training, completion
ReviewDecisions, corrections, review dates
04  Who we serve

Two markets. One accountability problem.

Organizations adopting AI

Mid-sized firms, associations, nonprofits and international implementers using AI without a mature internal governance function — where a board, a client, an insurer or a regulator has started asking questions nobody can currently answer.

Governments & development institutions

Institutions building policy capacity, procurement safeguards and durable oversight across African and cross-border markets — where international frameworks must be adapted, not imported.

05  Begin with visibility

Find out where your AI governance needs to start.

A short conversation about what your organization is running, who owns it, and what evidence exists. Scope and fee are confirmed after that conversation — never before.