AI governance discussions usually begin with models, agents, and automation.
But complex systems rarely fail because of the technology alone.
They fail because behavioral patterns shift faster than governance language can describe them.
Without stable terminology, those patterns remain invisible.
To address this problem, the Hollow House Institute maintains the AI Governance Glossary, a structured vocabulary for analyzing governance behavior in sociotechnical systems.
Version 1.3.0 now defines 41 canonical terms describing how governance dynamics emerge in human–AI environments.
Examples include:
• Behavioral Drift — gradual divergence between expected and observed system behavior.
• Authority Drift — shift in decision power away from intended governance structures.
• Override Erosion — decline in human willingness to interrupt automated decisions.
• Governance Illusion — appearance of governance control without effective enforcement.
These patterns often appear long before visible governance failures occur.
But without shared language to describe them, organizations struggle to detect them.
The purpose of the glossary is straightforward:
Create a common vocabulary for behavioral AI governance.
When governance language stabilizes, patterns become observable.
When patterns become observable, governance can move from aspiration to infrastructure.
Time turns behavior into infrastructure.
This post begins a short series introducing concepts from the AI Governance Glossary used to analyze governance dynamics in AI systems.
Canonical repository
[]https://github.com/hhidatasettechs-oss/Hollow_House_Standards_Library
Top comments (0)