An open-source model the team can self-host and inspect | AIGP
7-day money-back guarantee — full refund within 7 days of purchase if you've completed under 20% of the questions. See pricing →
Certifications Tools Flashcards Career Paths Exam Guides Blog Pricing For Teams About

Language

✓ EnglishDeutschEspañolFrançaisPortuguês
Check readiness — free →

An open-source model the team can self-host and inspect: Which profile best satisfies all three constraints?

AIGP Understanding How to Govern AI Deployment and Use Hard

An open-source, locally self-hosted model meets the inspection, air-gap, and data-residency constraints simultaneously.

The question

A defense contractor needs an AI model that must run inside an air-gapped facility, that its own engineers can fully inspect and modify, with no data leaving the premises. It is choosing among model-type profiles for the deployment. Which profile best satisfies all three constraints?

Preparing for AIGP? Take the free 5-min readiness quiz →

  1. An open-source model the team can self-host and inspect, deployed locally so that no data is transmitted beyond the air-gapped facility. ✓
    Correct because open-source weights allow full inspection and modification, and local self-hosting satisfies the air-gap and data-residency constraints together.
  2. A proprietary closed model accessed through the vendor's cloud API, chosen for its strong benchmark scores and managed reliability guarantees.
    Plausible for performance, but a cloud API breaks the air-gap and data-residency constraints and prevents full internal inspection.
  3. A proprietary closed model licensed as a black box for on-premise use, where the vendor withholds the weights and internal architecture.
    Plausible because it is on-premise, but withheld weights defeat the requirement that engineers fully inspect and modify the model.
  4. A hosted multimodal model that streams telemetry to the vendor for continuous improvement, chosen for its broad input-handling capabilities.
    Plausible for capability, but streaming telemetry externally violates the no-data-leaves-premises and air-gap constraints outright.
The trap
Assuming an on-premise deployment automatically grants the ability to inspect and modify the model.

How to remember it

An open-source, locally self-hosted model meets the inspection, air-gap, and data-residency constraints simultaneously.

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 — free

More 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.