Research & synthesis
AI may assist with structuring, summarising and exploring information. Material external claims should be checked against suitable evidence.
Responsible AI
AI Use Transparency Statement
Raeburn uses AI as an assistive capability where it improves analysis, engineering or productivity. AI does not remove human accountability, client confidentiality obligations or the need for evidence.
Where AI may assist
Drafting & analysis
AI may assist with first drafts, comparison, calculations, code and analytical work. Human responsibility is not transferred to the model.
Software engineering
AI may assist with implementation, tests, documentation and review. Changes remain subject to repository controls, testing and human accountability.
Internal productivity
AI may support routine administrative work where the data, permissions and consequence are appropriate.
Control boundaries
No autonomous professional authority
AI is not authorised to make legal, financial, regulatory, employment or other consequential commitments on behalf of a client or Raeburn merely because it generated a recommendation.
Confidentiality first
Client-confidential, personal or restricted data should only be processed in an approved environment and for an authorised purpose. Convenience does not override contract, privacy or confidentiality.
Human accountability
A named human remains responsible for material deliverables, recommendations and externally consequential actions.
Evidence over fluency
Plausible model output is not treated as evidence. Material facts, calculations and citations should be verified proportionately to consequence.
Tool permissions
Agent/tool access should follow least privilege, server-side authorisation and approval gates for high-impact actions.
Disclosure
Where AI use is material to the nature of an engagement or required by contract/regulation, the appropriate disclosure should be made rather than hidden.
Current disclosure
AI-assisted work may be used across research, drafting, analysis and software engineering. Material outputs remain subject to human review appropriate to their consequence. This statement describes governance principles; it does not claim that every task uses AI or that AI-generated output is independently correct.
Effective 27 August 2026 · Review on material change and at least annually.