Responsible AI
AI Principles
LinkAI is designed to support understanding and coordination while preserving authorised human judgement.
Version 0.1 · Reviewed 2 August 2026
People remain responsible
AI does not assume accountability for operational outcomes. Operational authority always remains with authorised customer personnel.
Advice, not instruction
LinkAI outputs are advisory and may be incomplete or incorrect. Users must assess them against current information, procedures and professional judgement.
Transparent and bounded
AI involvement and material limitations should be clear in context. Use cases are expected to have a defined purpose, accountable owner, proportionate evaluation, data controls, monitoring and a fallback.
- No autonomous operational control.
- No unapproved cross-customer or general-purpose training on Customer Data.
- No claim that AI is infallible or a substitute for competent judgement.
Public derivative ID: ai-principles. Internal source material remains controlled and is not incorporated into this page.
Need more detail?
Enterprise customers can request appropriate assurance information during due diligence. Controlled material is shared according to relevance and confidentiality.