Record each significant design decision with its rationale: Which documentation practice during design and
Documenting each significant decision with rationale, alternatives and owner creates the traceable, defensible record that establishes compliance and manages risk.
The question
Two years after building an AI system, regulators question why the team excluded a particular data source and chose a specific fairness threshold. Leadership realizes future disputes will hinge on being able to reconstruct such decisions. Which documentation practice during design and build best serves both compliance and risk management here?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Retain only the final trained model artifact and its headline accuracy score, on the basis that the deployed outcome is ultimately what regulators actually care about here.Plausible because the final artifact matters, but keeping only the outcome leaves no record of why decisions were made, so it cannot answer why a source was excluded.
- Keep informal chat logs and meeting invitations from the build period so the team can point to when the relevant discussions happened if questions later arise.Plausible because contemporaneous communications have some value, but scattered informal records do not systematically capture decisions and rationale in a defensible form.
- Record each significant design decision with its rationale, the alternatives weighed, and the responsible owner, so choices can be reconstructed and defended later. ✓Correct because documenting decisions, rationale, alternatives and ownership creates the traceable, defensible record that establishes compliance and manages risk when choices are later challenged.
- Preserve the raw training dataset in cold storage so the exact inputs remain available for any future re-examination of how the model came to behave.Plausible because retaining data aids reproducibility, but the dataset alone does not explain the reasoning behind excluding a source or setting a threshold.
The trap
Believing that retaining the final model, raw data, or informal communications documents design decisions, when only a structured decision log with rationale does. How to remember it
Documenting each significant decision with rationale, alternatives and owner creates the traceable, defensible record that establishes compliance and manages risk.
How many of these would you get right?
One of 1581 AIGP questions on Certsqill. Take a free five-minute check and see your score per domain — not one number, but which section to open tonight.
Test your AIGP readiness — freeMore Understanding How to Govern AI Development questions
- Remove the name feature from the design so the identified: Applying a risk mitigation hierarchy that prefers →
- It is not fit-for-purpose despite lawful rights and high: Under data-governance requirements, what is the most →
- Recording the origin of each data source and every: Which activity most directly fulfills that requirement? →
- All 426 Understanding How to Govern AI Development questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Development ·
Every answer, right and wrong, comes with its own explanation.
