part_of::[[Allen (2024) Progressive Trust]]
When a continuous framework (progressive trust) is operationalized as discrete stages (four authentication levels, six disclosure tiers), each stage boundary becomes a cliff — a point where the system’s behavior changes discontinuously. These cliffs need justification: why is this the boundary? What changes on either side?
IFP’s authentication level boundaries are justified by technical mechanism changes (symmetric → asymmetric crypto, self-asserted → externally resolvable, domain-specific → decentralized identity). But the disclosure tier boundaries are less clearly justified — why “professional-open” and “community-trust” as separate tiers? The progressive trust framework would ask: what verification happens at each boundary?
Transferable pattern: When discretizing a continuous model for protocol design, each stage boundary should be justified by a distinct verification event or trust mechanism change, not just a vocabulary label.
Progressive trust describes bilateral trust deepening between two parties. IFP adds a third dimension: personas. Trust is not just between Alice and Bob — it is between Alice-as-researcher and Bob, Alice-as-organizer and Bob, and potentially between Alice-as-researcher and Bob-as-clinician.
This means trust progression is not a single line but a matrix — each (persona, peer) pair has its own trust progression. The progressive trust framework needs extension to accommodate multi-context trust progression. This is a genuine contribution from IFP-12 back to the progressive trust theory.
Progressive trust says trust can be lost. IFP prevents authentication downgrade within an exchange but says nothing about between-exchange regression. In practice, trust regression happens: a person behaves badly, a key is compromised, a disclosure boundary is violated.
The mechanism for trust regression — how an agent records, communicates, and acts on lost trust — is a missing protocol in IFP. This may be a future IFP specification: “IFP-N: Trust Regression and Recovery.”
Transferable pattern: Any progressive trust implementation should specify not just how trust deepens, but how trust is lost, communicated, and potentially recovered.