It is not fit-for-purpose despite lawful rights and high: Under data-governance requirements, what is the most
Data governance requires fit-for-purpose alongside lawful rights and quality, so a dataset unrepresentative of the diverse deployment population is not fit for use.
The question
A team wants to train a medical-triage model using a large, lawfully licensed dataset of high image quality, but the images were collected from a single hospital serving a narrow patient demographic. The intended deployment spans a diverse national population. Under data-governance requirements, what is the most defensible position on using this dataset?
Preparing for AIGP? Take the free 5-min readiness quiz →
- It is acceptable to use because the data was lawfully licensed and is of high image quality, and those two attributes together satisfy the data-governance bar for training.Plausible because lawful rights and quality are genuine requirements, but they do not cover fit-for-purpose, so a narrowly representative dataset still fails the full governance test.
- It is acceptable to use because its large volume gives the model ample examples, and a sufficiently large sample will generalize to the broader population in practice.Plausible because volume aids learning, but size cannot compensate for a demographic gap, so a large but unrepresentative dataset remains unfit for the intended deployment.
- It is acceptable to use because the single-source origin makes lineage and provenance easy to document, and clean provenance is the decisive data-governance consideration here.Plausible because clean provenance is a real benefit, but easy lineage does not make unrepresentative data fit-for-purpose for a diverse deployment population.
- It is not fit-for-purpose despite lawful rights and high quality, because its narrow representativeness fails the deployment population and must be broadened before training. ✓Correct because data governance requires fit-for-purpose as well as lawful rights and quality, and a dataset unrepresentative of the deployment population fails that criterion regardless of licensing or image quality.
The trap
Assuming lawful rights, high quality, large volume, or clean provenance make data fit-for-purpose, when representativeness for the deployment population is a separate requirement. How to remember it
Data governance requires fit-for-purpose alongside lawful rights and quality, so a dataset unrepresentative of the diverse deployment population is not fit for use.
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
- Record each significant design decision with its rationale: Which documentation practice during design and →
- Recording the origin of each data source and every: Which activity most directly fulfills that requirement? →
- Examining whether model outcomes differ unjustifiably: Which purpose is served specifically by the bias →
- 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.