Provenance before persuasion
Material claims should be traceable to a source, direct evidence or clearly labelled analysis.
Responsible business & AI
Useful intelligence must also be accountable.
Our standards cover how evidence is collected, how analysis is communicated, how technology is used and where human judgement remains essential.
Operating principles
Promises that can be checked.
Proportionate verification
The significance of a decision determines the depth of checking, challenge and specialist review.
Visible uncertainty
Gaps, assumptions, confidence and competing interpretations are not hidden from decision-makers.
Human accountability
People remain responsible for recommendations, approvals and material consequences.
Responsible AI
AI may assist the work. It does not own the decision.
Permitted support
Structured research assistance, summarisation, drafting, classification, comparison and quality checks where inputs and outputs can be reviewed.
Human controls
A named person reviews material outputs, checks sources, challenges reasoning and decides whether the result is fit for purpose.
Restricted use
AI output is not treated as verified evidence, professional advice or autonomous authority for high-impact decisions.
Confidentiality: client or personal information should not be entered into unapproved public AI services. Data handling must follow the agreed purpose, access controls, retention terms and applicable privacy obligations.
Research governance
Every published intelligence item needs an owner and a review state.
| Control | Required record | Why it matters | Default review |
|---|---|---|---|
| Source record | Owner, title, URL or reference, period, retrieval date | Allows a reader to retrace material evidence | At publication |
| Analytical record | Author, assumptions, confidence, alternatives, limitations | Separates interpretation from fact | Before release |
| Content record | Publisher, publication date, last reviewed, next review trigger | Prevents undated authority claims | At least annually |
| Correction record | Issue, change, approval and correction date | Creates a visible route to remedy error | When identified |
Business conduct
Commercial judgement with clear boundaries.
Conflicts and independence
Relevant commercial interests or relationships should be disclosed where they may affect—or appear to affect—independent judgement.
Supplier and partner claims
A visit, introduction or relationship is not certification. Engagement-specific diligence and continuing verification remain necessary.
Outcomes and attribution
Client names, testimonials and quantified outcomes are published only where evidence and permission support the claim.
Accountability routes
Challenge the evidence. Report a concern. Request a correction.
Evidence
Case studies and provenance
See how direct activity, professional background and work in development are distinguished.
DecisionsDecision provenance
Explore the model for connecting evidence, review and accountable human judgement.
ContactRaise a question or correction
Contact Trixure and identify the page, claim or concern so it can be reviewed.
Trust is a practice, not a badge.
These commitments will evolve as Trixure’s services, products and evidence base develop.