AI Transparency Statement
This Statement explains, in plain language, what AI does inside RynIQ™, where humans intervene, and how Decision Confidence™ is composed. It is written to satisfy the transparency expectations of the EU AI Act (Articles 13 and 52) and the NIST AI RMF “Map” function without disclosing proprietary weightings or scoring formulas.
| Version | 2.0 | Effective | On publication |
|---|---|---|---|
| Supersedes | — | Governing law | Estonia · EU |
Purpose of the AI in RynIQ™
The AI components of RynIQ™ are used to structure and summarise executive input into a form that supports human reasoning. They are not used to make investment decisions, to select vendors, or to evaluate individuals.
How Decision Confidence™ is composed
Decision Confidence™ is a composite indicator expressed on a 0–100 scale. It combines four input families:
- Structured executive responses — the closed answers provided in the Executive Assessment.
- Governance indicators — the presence, quality and alignment of decision governance signals identified by the assessment.
- Investment readiness indicators — signals about problem clarity, sponsorship, and organisational capacity to execute.
- Human review — the reviewer's judgement, which can raise or lower the composite indicator.
The specific weightings, thresholds, and formulas that combine these inputs are proprietary. Publishing them would allow the score to be gamed and would reduce the value of the Executive Review to customers.
Decision Confidence™ is not a probability estimate, is not a recommendation, and is not a substitute for executive judgement. It is a structured signal used to focus the Review.
AI-authored content labels
- The Snapshot is entirely AI-drafted and labelled as preliminary.
- Sections of the Brief that are AI-drafted are marked so the reader knows.
- The reviewer's rewrite and additions form the reviewed portion of the Brief.
- Any output shared verbatim from a third-party model provider is treated as AI-drafted.
What the AI does not do
- It does not decide whether an investment should proceed.
- It does not evaluate vendors or negotiate commercial terms.
- It does not evaluate employees.
- It does not generate biometric, emotional, or protected-attribute inferences.
- It does not learn from customer data across engagements.
Model providers
We use enterprise-grade foundation model providers under contracts that prohibit training on our inputs and that apply zero-retention or short-retention terms. The current set of model providers is disclosed to enterprise customers on request under NDA.
Contact
Questions about AI transparency: privacy@ryniqo.com.