Create a documented data lineage record: Which missing control is most direct?
The unresolved issue is traceability: origins, transformations, and versions require a documented dataset lineage record.
The question
A media company is preparing an AI editing system. Its intended purpose, privacy basis, representative test set, acceptance criteria, subgroup results, and versioned model are documented. However, reviewers cannot establish where several training files originated or how they were transformed. Which missing control is most direct?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Expand the held-out test setA larger held-out set may improve evaluation coverage, but it does not establish origins or transformations of training files.
- Conduct additional subgroup performance analysisSubgroup analysis addresses differential performance, while the identified problem concerns undocumented source history and transformations.
- Create a documented data lineage record ✓A lineage record connects each dataset version to origins, transformations, and relevant handling decisions, resolving the stated evidence gap.
- Apply synthetic augmentation to missing filesSynthetic augmentation changes or supplements data but cannot establish the provenance of existing files with unknown origins.
The trap
When origins and transformations are unknown, select traceability documentation rather than another evaluation activity. How to remember it
The unresolved issue is traceability: origins, transformations, and versions require a documented dataset lineage record.
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
- Test plausible severe scenarios explicitly: Which testing distinction matters most? →
- Use a deployment-representative test sample: Which control is most direct? →
- Build an adjudicated: Which control most directly addresses both constraints before model training? →
- 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.