One number for how governed your AI really is.
A single, explainable 0–100 measure of your AI estate's health — five dimensions, qualified by confidence, tracked over time, traceable to every finding. Built to become the number your board asks about each quarter.
The Govern360 AI Exposure Score™ is a single 0–100 measure of how healthy and well-governed your AI estate is. Higher is stronger posture. It rolls up five operational dimensions, and every point traces to a specific, fixable finding — the difference between a number you trust and a black box.
The five dimensions
The score answers five executive questions, in a sequence you can recite: Discover → Govern → Protect → Control → Prove.
AI Discovery · 25%
Do you know every AI system in your environment? Inventory, shadow-AI detection, coverage. Discover →
AI Governance · 20%
Are AI systems governed consistently? Policy, approvals, ownership, oversight. Govern →
AI Protection · 25%
Is sensitive data protected? DLP prevention, AI identity, response scanning. Protect →
AI Control · 15%
Are costs, agents, and runtime under control? Token, budgets, MCP, agents. Control →
AI Compliance · 15%
Can you prove it? Continuous evidence across six frameworks. Prove →
How it is calculated
A weighted average of the five dimension scores, each computed from measurable facts — counts, ratios, coverage — not self-assessment. Weights are fixed within a model version and published, so a score change reflects your environment, not our math.
Always shown with the share of your estate that has been measured. An 82 at 20% coverage means something different than an 82 at 95% — so the score never hides its own confidence.
The change over time — +7 over 90 days, +14 since deployment. Momentum is what boards and renewals respond to, not a static grade.
Stale inputs cost points. If a source goes unscanned or evidence expires, the affected dimensions decay — always as an explainable finding with a cause and a fix, never silently — which keeps continuous governance honest.
Explainability is the point
A score that cannot be explained gets dismissed. The Govern360 AI Exposure Score™ drills all the way down, so every number is a worklist.
Score → Dimension → Module → Finding → Task. Each deduction is a finding with a point value and a remedy (“AI Protection 69 → extend DLP at point-of-use: +8 points”), and positive contributors show what is already working. The conversation shifts from “your score is 74” to “here is why, and here is how you get to 82.”
Questions, answered
What is the Govern360 AI Exposure Score?
It is a single 0–100 measure of how healthy and well-governed an organization's AI estate is. Higher is better. It decomposes into five operational dimensions — AI Discovery, Governance, Protection, Control, and Compliance — and every point traces to a specific finding you can act on.
How is the score calculated?
It is a weighted average of the five dimension scores (default weights: Discovery 25%, Protection 25%, Governance 20%, Control 15%, Compliance 15%). Each dimension is a rollup of module signals computed from measurable facts — counts, ratios, and coverage — not opinion. The weights are published on the methodology page.
What does the Confidence figure mean?
Confidence reflects how much of your estate has been connected and measured. An 82 at 20% coverage and an 82 at 95% coverage mean very different things, so the score is always shown with its Confidence and tells you what to connect next to raise it.
What is the Exposure Delta?
The change in your score over a period — for example, +7 over 90 days, or +14 since deployment. It turns the score from a snapshot into a momentum metric that boards and renewals respond to.
Why does my score decay?
Stale inputs cost points. If a source goes unscanned or evidence expires, the affected dimensions decay until the data is refreshed — which keeps the score honest and reinforces continuous governance. Every decay is shown as an explainable finding with its cause and fix, never silently.
Is the score explainable?
Yes — that is the point. Every deduction is a finding, and every finding is a task with a point value and a fix. The score answers not just “what is my number” but “why, and how do I improve it.”