Enterprise AI Measurement Guide
AI Governance · Policy Enforcement
Tools Triggering Enforcement
How many different AI tools are triggering policy violations, and are those violations concentrated in a few high-risk tools or spread across our entire shadow AI portfolio?
What it shows
Tools Triggering Enforcement shows how many distinct AI tools have generated at least one enforcement action (block or warning) during the selected period. Combined with the AI Tool Enforcement list, this count answers the breadth question: is enforcement activity concentrated on a handful of tools, or is it spread across the shadow AI portfolio?
Why it matters
A single tool generating 90% of enforcement events is a different governance problem than 30 tools each generating scattered enforcement activity. The first situation calls for a decision about one tool: evaluate it for approval or block it definitively. The second suggests the policy itself may need recalibration, either because the approved stack is not meeting employee needs, or because the classification logic is flagging too broadly. This metric is the triage view that tells leaders where to focus.
The Larridin angle
Larridin's enforcement layer is tool-aware rather than just event-aware, which means the same block action generated by Claude Desktop is distinguishable from one generated by an unknown third-party AI tool. That supports governance decisions that are proportionate to actual risk rather than treating all enforcement events as equivalent.