Glossary term
Evidence grade
Also: Evidence honesty layer, Source grading
A grade attached to each claim in a decision memo, showing whether the claim is backed by primary source, vendor-confirmed evidence, independent research, or an assumption. Assumption-heavy memos carry visible warnings rather than quiet confidence.
Memo quality often fails not at the conclusion, but at the evidence under each claim. A vendor-confirmed data point reads identically to a Slack-thread rumour unless the memo is structured to mark the difference.
A usable evidence grade scheme is small: primary (vendor documentation, signed contract language, certification artefact), independent (analyst report, security questionnaire, customer reference), internal (interview, architecture review) and assumed. A memo with more than 30 percent assumed claims is flagged, not published.
In practice
Evidence grading is the foundation of the anti-hallucination layer in DecisionOS. Every factual claim in a generated memo section must carry a grade; ungraded claims are blocked from the final output rather than smoothed over.
Related terms
Decision memo
A short structured document that captures why a decision was made, the options considered, the criteria and trade-offs, the stakeholders involved and the accepted risks. A good decision memo is auditable months later.
Readiness Score
A 0-100 score that quantifies how decision-ready a memo actually is. It breaks down into criteria coverage, evidence quality, risk analysis, stakeholder alignment and formal governance. A score above 70 signals audit readiness.
Defensible record
A single versioned artefact that captures a decision in enough structure, with enough evidence and stakeholder context, that it can be defended under audit, in a board review, or 12 months later when the original team is no longer in place.
EDR (Endpoint Detection and Response)
A class of endpoint security tools that continuously records endpoint activity and enables detection, investigation and response to threats that evade traditional antivirus. Most enterprise security incidents today are detected or contained at the EDR layer.
EU AI Act
World's first comprehensive AI law. Four risk classes: prohibited, high-risk, limited (transparency), minimal. Phased application from February 2025 to August 2028.
AI Act Conformity Assessment
Procedure to demonstrate that a high-risk AI system complies with the EU AI Act before being placed on the market or put into service. Mandatory for all Annex III high-risk systems from 2 December 2027 (postponed from the original 2 August 2026 by the Digital Omnibus, Regulation (EU) 2026/1744).
AI Act Risk Categories
Four-tier classification under the EU AI Act: prohibited (Art. 5), high-risk (Annex I/III), limited risk (Art. 50 transparency), minimal risk. Classification is the first duty; it must be documented and defensible.
Audit-ready decision
A decision whose record is structured, evidence-backed and stakeholder-signed to a level that a third-party auditor, a supervisor or a board reviewer can reconstruct the reasoning without interviewing the original team.