Canon term · Oversight

Kill Functionality

A capability, required to be maintained and exercisable rather than merely described, to immediately cease some or all automated action as an emergency measure — without the cooperation, consent, or continued functioning of the system it stops.

  • StatusIndustry term
  • AdoptionWidely used
  • LayerGovernance

Why it matters

Distinguish from Human Override, which is a person reversing one decision after the fact. Kill functionality is a pre-committed, system-wide capability that does not depend on a person noticing first — the whole point is that it fires whether or not anyone is watching at the moment it matters.

Where it comes from

Commission Delegated Regulation (EU) 2017/589 (RTS 6), applicable from 3 January 2018, requires investment firms engaged in algorithmic trading to maintain kill functionality permitting immediate cancellation of unexecuted orders. SEC Rule 15c3-5 (2010) requires the underlying controls to sit under the broker-dealer's direct and exclusive control, a closely related requirement predating RTS 6 by seven years.

Defining sources

  • Commission Delegated Regulation (EU) 2017/589 (RTS 6)European UnionOfficial text

What it is not

These are routinely confused with Kill Functionality. The distinctions have consequences for how a system is governed.

  • Human Override

    Override is a person reversing one decision. Kill functionality is a pre-committed capability to stop everything at once, without needing a person to notice first.

How it connects

What relies on it

MARQUE questions it addresses

  • DriftHow is behavioral change detected before it becomes loss?

Not to be confused with

Connected through the MARQUE (a reference framework organized around eight questions) crosswalk, an editorial mapping. These objectives from the Financial Services AI Risk Management Framework are assigned to the same questions that this term addresses. The ones sharing most of this term's vocabulary are shown first, as a reading aid. The assignment is to the question, not to the term: the source does not map this term to any objective directly.

Question: Drift · 11 control objectives assigned

  • MG-2.4.1Performance Monitoring and Action TriggersManage
All 11 objectives assigned to Drift

Cite and reuse

Cite the identifier QIS-TERM-00057, not the URL: identifiers are stable, URLs may change. This entry is free to read, quote and index under the dual license. Corrections sent to editor@qisfund.com are published.

Technical details: identifier, graph relationships and record
QIS-TERM-00057: outbound relationships as published in api/graph.json and api/lexicon.json
PredicateTarget identifierTargetPredicate meaning
addressesmarque:driftDriftThe subject speaks to the object as a question or concern.
relatedToterm:authority-envelopeAuthority EnvelopeAn association too weak or too general for a stronger verb.
notEquivalentToterm:human-overrideHuman OverrideThe two are routinely conflated and are distinct.
requiredByterm:restart-conditionRestart ConditionRead the other way round: the object requires the subject. The subject imposes the object as a precondition.
Canonical identifier
QIS-TERM-00057
Tier
Canonical industry term
Domain · layer
Oversight · Governance
Semantic aliases
kill criteria, kill switch criteria
First published
2026-08-02
Last reviewed
2026-08-02 · 180-day cycle