One Number You Can Defend
Aggregated Intelligence Posture for governance, risk, and compliance leaders.
Every organization adopting AI now runs at least three governance efforts at once. One tracks whether people use AI well. One tracks whether digital systems are ready for agents and partners. One tracks whether the organization meets its AI-specific legal obligations. Each produces its own dashboard. None of them produces the single answer a board actually asks for: how ready are you, on the whole, and can you prove it?
AI Posture answers that question with one level, bounded by the weakest of its in-scope vectors. It is an output measure. It scores verified behavior, not policies, intentions, or tool purchases. It is orthogonal to NIST AI RMF, ISO/IEC 42001, and EU AI Act conformance, which govern program design and remediation. This paper sets out the design rationale that makes a single readiness number defensible rather than reductive.
What is inside
The five design choices that make AI Posture defensible:
- A single readiness number across three independent vectors.
- An externally verifiable evidence standard.
- Aggregation by the weakest link rather than a flattering average.
- Treatment of the score as a time-stamped signal that decays rather than a certificate that persists.
- A clear boundary around what the framework does not claim to do.
Audio overview
A podcast-style conversation walking through the argument: Stop averaging your AI risks.
Video summary
A visual summary of the argument and the five design choices.