Human Oversight · Software component
Autonomy Scope Adjuster
Software componentHuman OversightExperience & Human Oversightarc:AutonomyScopeAdjuster
A feedback component that expands or narrows the agent's autonomous decision scope per case category by weighing demonstrated agreement with human decisions against the consequence of errors.
Responsibility. Recalibrates agent authority boundaries from observed performance and consequence.
Also known as: Authority boundary adjuster, Earned-autonomy controller
Relationships
reads dependency
writes dependency
receives data from dynamic
Design guidance
- MUST NOT adapt role boundaries purely on accuracy metrics without considering consequence.
Quantitative guidance
As stated by the sources; verify before use.
- Underwriting example: agent initially escalated 35% of self-employment claims; after six months it matched human judgment on 94%, justifying autonomous handling of standard cases (Ch10.2).
- 96% accuracy is unacceptable for high-value claims worth hundreds of thousands of dollars but may justify full automation for routine claims averaging $800 (Ch10.2).
Classification
- Patterns
- Earned autonomyConsequence-weighted authority
- Quality attributes
- Performance efficiency (ISO/IEC 25010)Safety (ISO/IEC 25010 | NIST AI RMF: safe)
- Risks mitigated
- Granting autonomy on accuracy alone for high-consequence decisionsRigid policies ignoring performance gains
Sources
- Ch10.2: T. Nguyen, "Proactive Agents," in Mastering Agentic AI Systems: Guide for the NVIDIA NCP-AAI Exam, 1st ed. 2026, ch. 10.2. ISBN: 9798244538229.