Assess and document the lawful rights to collect and use: Which action most directly addresses the threshold
For third-party scraped data, the threshold data-governance step is to assess and document lawful rights to collect and use it before ingestion.
The question
A data science team wants to train a model on a large corpus of images and text a vendor scraped from public websites. Before ingesting it, the AI governance function must apply data-governance requirements. Which action most directly addresses the threshold data-governance concern here?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Benchmark the trained model against a public leaderboard, since strong downstream accuracy demonstrates the data was of adequate quality and fit for the purpose.Leaderboard accuracy speaks to performance, not to lawful rights; a model can perform well on data the organization had no right to use.
- Encrypt the corpus at rest and restrict access to the team, since securing the data satisfies the core governance duty owed over collected training material.Security controls are necessary but do not resolve whether the organization has lawful rights to collect and use the data in the first place.
- Deduplicate records and normalize file formats first, since cleaning the corpus is what makes the dataset fit-for-purpose and therefore compliant to use.Cleaning improves quality and fit-for-purpose but does not address the lawful-rights question that scraped third-party data raises.
- Assess and document the lawful rights to collect and use the data, confirming the legal basis and permitted purposes before the corpus is ingested for training. ✓Establishing and documenting lawful rights to use the data is the foundational data-governance requirement, especially for scraped third-party content of uncertain provenance.
The trap
Believing that data quality, security, or performance can substitute for confirming lawful rights to use the data. How to remember it
For third-party scraped data, the threshold data-governance step is to assess and document lawful rights to collect and use it before ingestion.
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 design decisions: Which practice best fulfills that goal? →
- Documented data lineage and provenance tracing each: Which practice should have been in place to answer the →
- Bias testing to measure error-rate disparities across: Which pairing of test activities most directly →
- 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.
