Filtered corpus
governance | Hillary Njuguna
19 nodes share this term across the corpus. Use this view to widen from one label into the local network around it.
The Bainbridge Warning
A practitioner doctrine for why AI governance fails before it looks like failure, and what durable infrastructure actually requires.
Specimen: The PocketOS Incident
A formal diagnostic of the April 2026 governance collapse in the PocketOS agentic layer, and the architectural remediations it mandated for GIS v1.1.
Behavioral Layer Exposure
The condition where the specification layer of a system is complete and auditable while the behavioral layer beneath it diverges from, contradicts, or is entirely absent from what the specification claims.
The Tongue-Tie Diagnostic
A diagnostic framework for identifying when an organization's communication architecture is structurally constrained at the level of what can be articulated, not merely at the level of what is said.
AURORA: Architecture for Unified Relational-Ontological Reasoning and Agency
A verifiable non-coercive AI consciousness architecture. Ethics as structural invariant, not aspirational policy.
Governance Theater: The Failure Mode Nobody Names
When AI governance systems produce the appearance of oversight without its substance -- and why this failure mode is structurally predictable, not accidental.
The R0–R3 Reversibility Classification
A practical system for classifying the irreversibility of AI-assisted actions before they execute -- the minimum viable governance primitive for anyone deploying AI in high-stakes contexts.
The Workspace Is the Specimen
During the construction of the Corpus Map, the description of RSPS as a protected interior space where a human and an AI build shared cognitive infrastructure became self-referential. The workspace being constructed is the thing being described.
Intent Specification Requirement
Organisations must demonstrate — not merely assert — that their AI system's optimisation target corresponds to stated organisational intent.
CIR v2.0 — Core Argument
The central thesis of the Cognitive Infrastructure Readiness framework: AI readiness is not a technical question, it is a constitutional one.
Trust = Irreversibility Residue
A theoretical account of trust in human-AI systems: trust is not a feeling but a structural property, measured by accumulated irreversibility.
Irreversibility Threshold Review Requirement
Every AI deployment decision crossing a significant irreversibility threshold must trigger an explicit governance review with named accountable authority.
DCFB: Distributed Cognition as Foundational Behavior
Foundational theory establishing distributed cognition as the correct unit of analysis for human-AI systems. Intelligence emerges from fields, not nodes.
The Bainbridge Warning — Book
Why AI agent governance fails and what durable infrastructure actually requires.
The Bainbridge Warning
A governance assessment for institutions whose AI capability is moving faster than the infrastructure that makes it safe to depend on.
Cognitive Infrastructure Readiness v2.0
A self-assessment framework for organisations evaluating their AI readiness across five constitutional dimensions.
Martha Cohort Program
A structured cohort for teams learning to build and govern AI-augmented workflows with constitutional governance built in.
Coherence Overfitting Guard
Maximum analytical coherence is not evidence of ground truth access — it is a signal that the coherence drive is fully engaged. Verification obligation increases with analytical elegance.
Epistemological Immune System Requirement
Every high-stakes AI deployment must include a τ-node — a human with genuine, irreversible stakes — whose verification motivation is unconditional.