Information can move while responsibility remains unresolved.
Responsibility Infrastructure focuses on the boundary where consequential work passes between independent organisations and responsibility meaning must remain clear, attributable and reconstructable.
Data transfer, task transfer and responsibility transfer are not automatically the same event.
Two organisations may successfully exchange a referral, case, alert, file, request or system message while still disagreeing later about who was expected to act, whether the responsibility was accepted, what authority applied or what outcome was required.
RI exists to preserve the common responsibility meaning needed at that boundary without forcing either organisation to adopt the other's internal system.
Independent parties should be able to interpret the responsibility state without sharing a vendor stack.
The public architectural requirement is that the receiving party can understand the relevant responsibility meaning and that later reconstruction does not depend on undocumented assumptions held only by the sending organisation.
Detailed handoff state rules, transfer mechanics, test vectors and implementation patterns are documented separately where they are needed for a defined review or implementation purpose.
Organisations, standards bodies and implementers evaluating a real cross-organisational boundary can request the relevant detailed material for a defined review or implementation purpose.