Use procurement assessments and contract terms setting: Which measure most directly manages this third-party
Third-party risk is best managed through vendor due-diligence assessments and contract terms that set obligations and allocate liability before integration.
The question
A company plans to embed a third-party vendor's large language model into a customer-facing product. It cannot inspect the vendor's training data and worries about liability if the model produces harmful outputs. Which measure most directly manages this third-party risk?
Preparing for AIGP? Take the free 5-min readiness quiz →
- Rely on an internal acceptable-use policy instructing employees on how to responsibly interact with the deployed model throughout their daily work.Plausible and useful, but an internal usage policy does not bind the vendor or allocate liability for the model's outputs.
- Expand the internal AI literacy program so staff understand the model's general capabilities, key limitations, and appropriate use cases.Almost right, since awareness helps, but training employees does not address vendor accountability or the contractual liability concern.
- Increase internal model monitoring dashboards so the product team can observe the system's live output patterns after full deployment.Plausible as an operational control, but post-deployment monitoring alone does not manage the pre-contract vendor risk and liability exposure.
- Use procurement assessments and contract terms setting vendor obligations, representations, and liability allocation before integrating the model. ✓Correct. Vendor due-diligence assessments plus contractual obligations and liability allocation are the primary levers for third-party AI risk.
The trap
Believing internal policies and monitoring can substitute for the vendor assessments and contractual terms that actually govern third-party risk. How to remember it
Third-party risk is best managed through vendor due-diligence assessments and contract terms that set obligations and allocate liability before integration.
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 the Foundations of AI Governance questions
- Intellectual property and data governance policies: Which existing policy area most urgently needs updating →
- A system that learns patterns from data to make: Which description best characterizes a machine learning →
- Ethics and bias risk: Which category of AI risk does this best illustrate? →
- All 337 Understanding the Foundations of AI Governance questions →
Part of the Certsqill AIGP question bank · Understanding the Foundations of AI Governance ·
Every answer, right and wrong, comes with its own explanation.
