Bounded Autonomy: Behavioral Specification Languages and Runtime Enforcement Architectures for Trustworthy Agentic AI Systems Authors Harper G
Innovative Journal of Applied Science · 2026 · 03 (02) · 01-07
DOI: 10.70844/ijas.2026.3.50 ↗Kurz gefasst
Fachartikel in der Zeitschrift Innovative Journal of Applied Science, 2026. Verfasst von Harper Gough.
Abstract (Originalsprache)
Agentic AI systems autonomously plan and execute multi-step workflows using tools, external services and mutable environments. This shifts the engineering risk from single-call model correctness to end-to-end behavioral predictability across time, resources, dependencies and adversarial conditions. Current agent research largely optimizes capability, whereas the reliability and trustworthiness of agent behavior remain weakly specified, monitored and enforced at runtime. Existing AI safety work provides limited guidance for engineering trustworthy agents at current capability levels and classical software specifications and runtime verification methods do not transfer directly because agent behavior emerges from learned models interacting with tools and context rather than explicit code paths. This study frames the central deployment problem as bounded autonomy: preserving beneficial agent initiative while enforcing a behavioral envelope that is auditable, resilient and secure. This study proposes a six-dimensional human-centered AI framework adapted to agentic systems and uses it to identify gaps in behavioral specification languages, runtime enforcement architectures, competence-boundary handling and graceful degradation. It then defines evaluation metrics and methodology grounded in fault injection, resilience testing, adversarial ML and production-readiness practices and provides case analyses in safety investigation operations, infrastructure management and cyber-physical automation.
Zitieren
APA
Gough, H. (2026). Bounded Autonomy: Behavioral Specification Languages and Runtime Enforcement Architectures for Trustworthy Agentic AI Systems Authors Harper G. Innovative Journal of Applied Science, 03(02), 01-07. https://doi.org/10.70844/ijas.2026.3.50
Vancouver
Gough H. Bounded Autonomy: Behavioral Specification Languages and Runtime Enforcement Architectures for Trustworthy Agentic AI Systems Authors Harper G. Innovative Journal of Applied Science. 2026;03(02):01-07. doi:10.70844/ijas.2026.3.50
BibTeX
@article{Gough2026,
author = {Gough, Harper},
title = {Bounded Autonomy: Behavioral Specification Languages and Runtime Enforcement Architectures for Trustworthy Agentic AI Systems Authors Harper G},
journal = {Innovative Journal of Applied Science},
year = {2026},
volume = {03},
number = {02},
pages = {01--07},
doi = {10.70844/ijas.2026.3.50}
}
EUVIMED