The COMPEL Glossary Graph visualizes relationships between framework terminology, showing how concepts interconnect across domains, stages, and pillars. Term nodes cluster by pillar affiliation while cross-references reveal semantic dependencies — for example, how risk appetite connects to control effectiveness, model governance, and assurance requirements. This network representation helps practitioners navigate the framework vocabulary and understand that COMPEL terminology forms a coherent conceptual system rather than isolated definitions.
COMPEL Glossary / runtime-intervention
Runtime intervention
An oversight mode in which a human can redirect, pause, or halt an agent during execution — through a dashboard, approval prompt, or interrupt signal — without requiring pre-authorization for every action.
What this means in practice
Maps to EU AI Act Article 14(4)(b) oversight-measure category; appropriate for agents with moderate autonomy and reversible side effects.
Synonyms
in-flight intervention , human-on-the-loop intervention
See also
- Pre-authorization oversight — An oversight mode in which a named human must approve specified agent actions before they execute.
- Post-hoc review — An oversight mode in which human reviewers inspect agent actions after they complete, typically on a sample, exception, or threshold basis.
- Stop-go decision right — The explicit assignment of authority to halt or resume an agent — distinct from authority to configure, deploy, or operate it.
- Human oversight (Art. 14) — Under Regulation (EU) 2024/1689, the provider-designed measures that allow natural persons to understand the capacities and limitations of a high-risk AI system, monitor its operation, and intervene or interrupt it.