Definitions of the named frameworks and core terms used across this site — for the reader (or the AI system) that wants the precise meaning, not just the argument.
Proprietary Frameworks
From The Verification Economy
The Verification Economy™
The operating model for a market where AI has made generation free — and verification, not output, is the new competitive advantage. Read the full piece →
Cost-per-Validated-Asset™ (CVA)
A formula that makes the hidden cost of AI verification visible and measurable, exposing what standard productivity dashboards conceal. Read the full piece →
The Three Control Taxonomies
HITL, HILE, and HOTL — a framework for assigning the right level of human oversight to each AI workflow, based on risk and velocity. Read the full piece →
The Verification Maturity Model™
A four-stage path from ad hoc AI adoption to "The Trust Factory" — where governance itself becomes a priced commercial asset. Read the full piece →
Safe AI SME Series
From the Book Companion Articles
Workload Creep
The phenomenon where AI-enabled speed silently triples output expectations without any corresponding reduction in the cognitive demand of doing the work well. Read the full piece →
Prompt Churn
The cycle of entering a vague prompt, generating, reviewing, adjusting, and regenerating — often eight to fifteen times — which quietly consumes the time AI was meant to save. Read the full piece →
Verification Fatigue
The psychological shift from builder to auditor that happens when reviewing AI-generated output, and the decline in review quality that compounds silently across weeks. Read the full piece →
Compliance Bias
The point at which a fatigued reviewer — worn down by output that's usually correct — begins to skim, assume, and approve without real scrutiny. Read the full piece →
Total Task Time
The honest metric for AI productivity: prompt formulation time, plus generation time, plus human verification and editing time, plus approval time — compared against the manual baseline, not just the generation step. Read the full piece →
The Open Tab
Shorthand for the unmanaged, consumption-based token-billing risk of agentic AI — an autonomous pipeline that can run up a five-figure invoice overnight with no human watching. Read the full piece →
Shadow AI
Unmanaged or unsanctioned use of free consumer AI tools inside a business, without a governance structure, data processing agreement, or audit trail. Read the full piece →
General Terms
AI & ESG Governance
AI Governance
The systems, policies, and oversight structures that make AI decision-making traceable, auditable, and accountable to regulators, boards, and clients.
ESG Governance
The frameworks and disclosure practices that ensure environmental, social, and governance claims are accurate, evidenced, and defensible under audit.
Algorithmic Ethics
The practice of ensuring AI and machine learning models treat stakeholders fairly and do not encode or amplify bias against protected groups.
Human-in-the-Loop (HITL)
A governance model requiring human review and approval of an AI-generated output before it takes effect or reaches a customer, regulator, or investor.
Readiness Assessment
A structured diagnostic of an organisation's AI and ESG governance maturity, data integrity, and regulatory exposure, producing a sequenced roadmap aligned to commercial priorities.