1. Traceable evidence
Important numbers, claims and explanations remain connected to supporting information. Missing, unclear or conflicting evidence is made visible rather than silently filled in.
2. Controlled quantitative work
Where material calculations are required, they use reproducible methods appropriate to the work and remain connected to the relevant source information.
3. Visible assumptions, uncertainty and exceptions
Assumptions, unresolved questions, missing evidence, conflicting information and items requiring professional judgment remain explicit.
4. Accountable review and judgment
AI can help organize evidence, analyse information, identify patterns, construct narrative and challenge interpretations. Accountable professionals retain responsibility for materiality, interpretation, review, judgment and decisions.
5. Continuity where prior context matters
Where prior context affects a later cycle or case, relevant mappings, decisions, explanations, corrections, assumptions, reviewer judgments and open exceptions can be carried forward and kept current — corrected, superseded or retired as circumstances change.