Behavioral AI Governance is the study of how artificial intelligence systems evolve through repeated operational interaction with humans, workflows, and institutional processes. It focuses on execution-time behavior rather than only policy documentation.
Governance Drift (HHI-GOV-013) describes the divergence between documented governance policies and the actual behavior of deployed AI systems over time.
Execution-Time Governance describes governance mechanisms that operate during system operation, ensuring human authority, accountability, and decision oversight while AI systems are actively being used.
Behavioral AI Governance Research | AI Governance Terminology | Execution-Time Governance
The Hollow House Institute (HHI) develops governance terminology, frameworks, and research infrastructure for understanding how artificial intelligence systems behave in operational environments.
Time turns behavior into infrastructure.
Behavior is the most honest data there is.
Behavioral AI Governance studies how artificial intelligence systems evolve through repeated operational interaction with humans, workflows, and institutions.
Rather than focusing only on policy or compliance documentation, this approach examines execution-time system behavior.
Key dynamics include:
- Behavioral Drift (HHI-BEH-001)
- Decision Substitution (HHI-AUTH-004)
- Override Erosion (HHI-BEH-004)
- Governance Drift (HHI-GOV-013)
- Governance Infrastructure (HHI-GOV-002)
HHI Governance Glossary
https://github.com/hhidatasettechs-oss/Hollow_House_Standards_Library
This glossary functions as a machine-readable terminology registry defining core concepts used in Behavioral AI Governance research.
HHI_GOV_01 — Execution-Time Governance Standard
https://github.com/hhidatasettechs-oss/HHI_GOV_01
This governance standard describes mechanisms that preserve human decision authority during AI system operation.
Audit repository:
https://github.com/hhidatasettechs-oss/HHIaudits
These artifacts document real governance dynamics observed in deployed AI systems.
Zenodo DOI
https://doi.org/10.5281/zenodo.18615600
LinkedIn
https://www.linkedin.com/in/hollow-house-institute
DEV Community
https://dev.to/hollowhouse
Governance must precede agent autonomy.
