The Reed Group
Governance Instrument

AI Readiness
Diagnostic

A structured assessment of your organization's AI governance posture across the Four Signals Framework: Ownership, Exposure, Authority, and Response.

This diagnostic is not a technology audit. It does not measure whether your AI systems perform well. It measures whether your organization can govern them.

Answer each question based on the current documented reality of your organization, not the policy you intend to write. Honest responses produce useful findings. Optimistic responses produce a false tier rating and no practical value.

When you complete all 20 questions, you receive a one-page Signal Map: a risk tier rating across all four governance dimensions, with dimension-specific findings and a prioritized action for each. You leave this instrument knowing exactly where your governance posture stands and what to address first.

If you are a Board Member or C-Suite Executive
Use this to understand what your management team should be able to document. The four signals map directly to the governance questions your board will ask this year.
If you are a CISO, CSO, or Compliance Leader
Use this as a structured pre-assessment before a Governance Diagnostic engagement, or as a rapid self-audit against the 2026 benchmark. Answer based on documented reality, not intended policy.
0 of 20 questions answered
Live Tier Reading — Four Signals
Ownership
-
Exposure
-
Authority
-
Response
-
Overall Governance Tier Incomplete
01
Signal One
Ownership
Who is accountable for each AI system in production, and do they know it?
O-1 — InventoryDoes your organization maintain a documented inventory of every AI system currently in production or actively in use by employees?
O-2 — Named AccountabilityFor each AI system in use, is there a named individual (not a team, not a department) who is formally accountable for its governance?
O-3 — Shadow AIHas your organization assessed whether employees are using AI tools not sanctioned or reviewed by IT, legal, or security?
O-4 — Vendor HandoffsWhen a vendor deploys or hands over an AI-enabled system, does your organization have a formal process for accepting governance responsibility?
O-5 — Board VisibilityCan your CISO or CTO produce a complete AI ownership picture if asked by the board today, without a preparation period?
02
Signal Two
Exposure
What data, systems, and decision points are touched by AI, and who can see them?
E-1 — Data ClassificationHas your organization mapped which categories of data are being ingested, processed, or referenced by AI systems currently in use?
E-2 — Third-Party Data FlowDo you know whether employee use of AI tools results in proprietary or sensitive data being transmitted to third-party model providers?
E-3 — Access ControlsAre access controls in place restricting which employees can interact with or configure AI systems that touch sensitive data or critical operations?
E-4 — IP ExposureHas a legal or security review assessed whether current AI usage patterns create intellectual property exposure risk?
E-5 — Audit TrailDoes your organization maintain logs of AI system activity sufficient to reconstruct what actions were taken or what data was accessed during an incident?
03
Signal Three
Authority
Where does the AI decide, where does the human decide, and is that boundary written down anywhere?
A-1 — Decision Boundary DocumentationFor AI systems used in consequential decisions, is there documented policy specifying where AI judgment ends and human judgment is required?
A-2 — Human-in-the-LoopIs there a formal human review requirement before AI-generated outputs are used to make binding decisions in high-stakes contexts?
A-3 — Regulatory AlignmentHas legal or compliance reviewed your AI decision practices against applicable regulations (EU AI Act, state AI laws, sector-specific requirements)?
A-4 — Override ProtocolDo employees have a clear, documented process for overriding an AI recommendation when their judgment differs?
A-5 — Governance PolicyDoes a formal AI governance policy exist that has been reviewed by leadership and communicated to employees?
04
Signal Four
Response
If an AI system fails in a way that causes harm, what happens in the first four hours?
R-1 — Incident DefinitionHas your organization formally defined what constitutes an AI incident, as distinct from a conventional IT or security incident?
R-2 — Response PlaybookDoes a documented AI incident response playbook exist specifying who is notified, in what order, and what actions are taken in the first four hours of a governance failure?
R-3 — Escalation PathIf an AI system produced an output that caused material harm, is there a clear escalation path to the board or senior leadership within a defined time frame?
R-4 — Shutdown AuthorityIs there a documented authority chain for suspending or shutting down an AI system if it is producing harmful or anomalous outputs?
R-5 — Board ReadinessCould your CISO or legal counsel provide a credible account of an AI governance failure to your board within 24 hours, including cause, scope, and initial remediation steps?
Your Signal Map
Next Step
This is where the conversation starts.
Request a Diagnostic Conversation