Responsibility Attribution
Responsibility attribution is the assignment of accountability to specific actors at the point where reasoning becomes consequential.
The Attribution Problem
In complex systems, outcomes are often visible, but responsibility is not.
Multiple actors may contribute to a decision:
- humans
- AI systems
- institutions
Without clear attribution, accountability becomes diffuse.
From Contribution to Responsibility
Not all participation implies responsibility.
Responsibility arises at the point of:
- selection
- validation
- promotion
This typically occurs at a responsibility boundary.
Relation to Promotion Gate
A promotion gate is where reasoning becomes durable.
Responsibility attribution identifies who is accountable at that transition.
Role of HuLoo
In PKOS, attribution is often made explicit through HuLoo checkpoints.
At these points, humans:
- interpret reasoning
- make selections
- assume responsibility
Attribution is not inferred — it is declared.
Failure Mode
Without responsibility attribution:
- decisions appear without owners
- accountability becomes ambiguous
- errors cannot be traced or corrected effectively
This is a defining characteristic of Shadow AI.
Relation to Decision Memory
Decision memory preserves the context of a decision.
Responsibility attribution ensures that the decision is linked to an accountable actor.
Governance Role
Responsibility attribution enables:
It connects decisions to responsibility, not just to reasoning.
Summary
Responsibility attribution answers:
- who is accountable?
- at what moment?
- for which decision?
It transforms reasoning into accountable action.
Part of the PKOS Lexicon.