AI Risk Management Framework 1.0
Organizes AI risk management around Govern, Map, Measure, and Manage across the lifecycle.
NIST — official/source page ↗A layered research architecture that keeps rights, jurisdiction, purpose, epistemic type, evidence, authority, appeal and restoration above individual machine models.
It is a technical-governance proposal, not executable constitutional law: a 13-layer architecture designed so that a model cannot grant itself authority. Rights and jurisdiction sit upstream of inference; appeal, restoration and audit sit downstream of action.
The stack begins with fundamental human rights, machine jurisdiction, permitted purpose and data authority. Only then does it reach epistemic type, evidence standard, model inference, decision authority and response authority. Appeal, restoration, retention and audit complete the lifecycle.
An accurate output can still be jurisdictionally invalid, purpose-invalid, procedurally invalid or disproportionate. Constitutional correctness therefore cannot be reduced to predictive accuracy.
Permissions, purpose tags, retention, version lineage and sunset conditions can be machine-tested. Human Standing, proportionality and democratic legitimacy cannot be fully reduced to code.
Source links establish traceability and support. They do not imply that the source endorses Concresca’s constitutional proposals.
Organizes AI risk management around Govern, Map, Measure, and Manage across the lifecycle.
NIST — official/source page ↗Includes separation-of-duties and least-privilege controls relevant by analogy to separation of machine powers.
NIST — official/source page ↗Defines a W3C Recommendation for representing provenance entities, activities, and agents.
W3C — official/source page ↗Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
Open the v0.14 implementation framework.
NO JUDGMENT WHATSOEVER. Concresca coordinates without assigning moral worth, character, guilt, danger, trustworthiness, loyalty, purity, normality, or social standing. Questions, thoughts, identities, messages, content, and conduct are not objects of Concresca judgment.