🟢 Continuity of Agency
Continuity of Agency is a human's unbroken understanding of a system's actual state, sustained across time — not a single act of oversight, but oversight that doesn't lapse between one decision and the next.
Not the Same as Human Agency
Human Agency establishes that a person remains the point of decision. Continuity of Agency is the temporal condition underneath that: the same person's grasp of what the system is actually doing has to stay current, or the decision they're making is about a system that no longer exists in the form they last understood.
Where It Breaks
Agency doesn't usually disappear all at once. It erodes quietly, one unreviewed change at a time, until the person nominally in charge is deciding based on a mental model of the system that's now several versions out of date. Nobody took the decision away from them. The system simply moved faster than their understanding of it did.
This is a slower, less visible cousin of Shadow AI — not reasoning disappearing from the institution's view, but the specific human's own comprehension quietly falling out of step with a system they're still, formally, responsible for.
Why It Needs Its Own Name
A single well-designed Promotion Gate checkpoint can be passed correctly and still not be enough, if the person passing it hasn't kept up with everything that changed since the last time they looked. Continuity of Agency is what makes each individual checkpoint actually meaningful rather than a formality — the person at the gate has to still understand the system standing in front of them.
Continuity of Agency is Human Agency's own temporal half: Navigational Coherence's insistence that correction stays possible, applied specifically to the person who's supposed to be doing the correcting.
Part of the PKOS Lexicon.