Assess the retailer as a potential provider: What does this evidence most defensibly support?
Substantial modification, changed purpose, and own-name marketing support examining potential provider responsibilities, not relying on purchase status or accuracy.
The question
In the EU AI Act context, a retailer substantially retrains a vendor’s demand-planning system on proprietary data, changes its intended purpose from forecasting to automated inventory allocation, and markets the modified tool under its own name. Evidence shows improved accuracy but no analysis of the role change. What does this evidence most defensibly support?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Treat the retailer solely as a deployer because it purchased the original system.Purchase alone does not determine role; substantial modification and changed intended purpose can create additional provider responsibilities.
- Assess the retailer as a potential provider. ✓The substantial modification, changed intended purpose, and own-name marketing may shift provider responsibilities under Article 25 conditions.
- Classify the system as low-risk because inventory allocation is commercial.Risk classification depends on intended purpose and applicable categories, not merely on the commercial setting or improved accuracy.
- Accept vendor accountability because accuracy improved after retraining.Performance improvement does not resolve whether modification, intended purpose, and marketing altered the retailer’s statutory role.
The trap
Separate technical improvement from legal-role analysis; examine modification, intended purpose, and market presentation together. How to remember it
Substantial modification, changed purpose, and own-name marketing support examining potential provider responsibilities, not relying on purchase status or accuracy.
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 Laws, Standards and Frameworks Apply to AI questions
- Assess whether modification and own-name distribution: What does the evidence support? →
- Document mitigations and post-deployment verification: Which documentation action best addresses the evidence →
- Assess representativeness and label quality across: What evidence should it prioritize? →
- All 394 Understanding How Laws, Standards and Frameworks Apply to AI questions →
Part of the Certsqill AIGP question bank · Understanding How Laws, Standards and Frameworks Apply to AI ·
Every answer, right and wrong, comes with its own explanation.
