Treating component performance as proof of end-to-end value.
Research evaluation lens
System boundary for “Consent, selective disclosure and provenance”
Define the unit of analysis, operating context and dependencies before interpreting any result.Independent depth pageEvidence developingReviewed 12 August 2026
Purpose & scope
What this lens examines—and what it does not prove.
For “Consent, selective disclosure and provenance”, the system is larger than a model or interface. It includes the people who frame and review the work, the tools and data it depends on, the organisation that authorises action, and the physical or institutional environment in which outcomes occur.
A boundary is useful only when exclusions are explicit. It should identify which users, tasks, jurisdictions, time horizons and failure conditions are outside the present claim, so that a local result is not mistaken for a universal one.
Evidence state
Developing: this page organises public evidence and evaluation questions; it does not claim that every condition has already been tested.
Diagnostic questions
Questions an evaluation should be able to answer.
Answers should identify evidence, owners and conditions—not only intentions.
- 01
What is the exact unit being evaluated: a model, workflow, team, firm, market or public system?
- 02
Which people can authorise, override, inspect or stop the system, and when are they expected to act?
- 03
Which data, tools, vendors, institutions and physical conditions are necessary for the result?
- 04
Which populations, tasks, locations and time periods are deliberately outside the claim?
- 05
What changes in the surrounding environment would invalidate the present boundary?
Evidence plan
Records needed before the lens can support a decision.
Evidence should remain attributable and preserve uncertainty, counterexamples and context.
- 01
A system map naming actors, interfaces, dependencies and control points.
- 02
Task and context definitions that can be reproduced by an independent reviewer.
- 03
Records of exceptions, hand-offs and conditions in which human judgement replaces automation.
- 04
A written exclusion list and transfer tests before applying results to a new setting.
Public evidence anchors
Trace the frame back to attributable sources.
Sources inform the frame; inclusion does not imply collaboration, review or endorsement.
01W3C
Verifiable Credentials 2.0
An interoperable, privacy-aware foundation for machine-verifiable claims.
Open canonical source ↗02NISTAI Agent Security and Identity
Public standards work connecting identity, authorisation and secure adoption.
Open canonical source ↗03FUURAA Frontier LibraryMemory, credentials and provenance signals
A bilingual evidence trail maintained from checked primary sources.
Open canonical source ↗FUURAA analysis
Use the lens to improve a decision, not decorate a claim.
FUURAA’s analysis is that “Consent, selective disclosure and provenance” should be evaluated at the smallest boundary that still contains the real decision, its consequences and the people accountable for them. A narrower boundary may make a metric look cleaner while excluding the labour, risk or infrastructure that determines whether the result has practical value.
Applicable boundaries
- This page is a research framework, not a product capability or deployment claim.
- The correct boundary depends on the decision being made and may change as the system changes.
- A clear boundary improves interpretation but does not by itself establish effectiveness or safety.
Decision-ready outputs
- A bounded claim that states where the finding applies.
- An ownership map for review, intervention and escalation.
- A list of boundary changes that trigger re-evaluation.
Parent research brief