On-premise deployment: Which deployment option most directly satisfies that constraint?
On-premise or edge deployment keeps sensitive data and inference within the organization's own infrastructure.
The question
An organization must deploy an AI system in a way that keeps highly sensitive data from ever leaving its own controlled infrastructure. Which deployment option most directly satisfies that constraint?
Preparing for AIGP? Take the free 5-min readiness quiz →
- A public cloud API from a third-party provider, which sends each request to the vendor's external environment for processing and for its own logging.Wrong because routing requests to a third-party cloud sends the sensitive data outside the organization's boundary.
- On-premise (or edge) deployment, which keeps sensitive data and model execution within the organization's own controlled infrastructure boundary at all times. ✓Correct because on-premise or edge hosting keeps data and inference inside the organization's own environment.
- Retrieval augmented generation, which improves answer relevance but does not by itself determine where the underlying data is physically processed or stored.Plausible but wrong because RAG is a technique for grounding answers, not a control over data residency.
- Fine-tuning the model on additional in-house data, which changes the model's behavior but does not by itself control where inference ultimately takes place at run time.Plausible but wrong because fine-tuning adjusts behavior without controlling where data is processed.
The trap
Confusing modeling techniques (RAG, fine-tuning) with the hosting choice that actually controls data residency. How to remember it
On-premise or edge deployment keeps sensitive data and inference within the organization's own infrastructure.
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 Deployment and Use questions
- A classic, more deterministic-leaning model tends: Which consideration about model type is most relevant to →
- Before deployment and refreshed when material changes: When should the impact assessment be performed relative →
- Rights to necessary technical documentation: Which combination of terms best protects it? →
- All 424 Understanding How to Govern AI Deployment and Use questions →
Part of the Certsqill AIGP question bank · Understanding How to Govern AI Deployment and Use ·
Every answer, right and wrong, comes with its own explanation.
