Stakeholder mapping to find who could be harmed: Which combination best reflects sound pre-deployment risk

Stakeholder mapping plus a limited pilot under realistic conditions are structured methods that surface risks before full deployment.

The question

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

1. A company-wide launch to gather the most feedback quickly, accepting that some early users may encounter serious, unmitigated harm along the way.

   Plausible for fast learning, but a broad launch exposes many users to unvetted risk rather than containing it.
2. Stakeholder mapping to find who could be harmed, combined with a limited pilot that tests the system under realistic operating conditions. ✓

   Correct: mapping affected parties and running a contained pilot are recognized structured methods for surfacing risks before full deployment.
3. Reliance on the vendor's marketing benchmarks, treating the published accuracy claims as sufficient evidence that residual risk is low.

   Plausible as convenient, but vendor claims are not independent risk identification.
4. A single sign-off from the project sponsor, whose approval is taken as confirmation that the residual risks are acceptable.

   Plausible as governance, but one approval does not identify or evaluate the risks themselves.

How to remember it

Stakeholder mapping plus a limited pilot under realistic conditions are structured methods that surface risks before full deployment.

More Understanding How to Govern AI Development questions

• Engaging affected stakeholders and setting metric: Which practice best reflects embedding ethics into the →
• Record design decisions: Which practice best fulfills that goal? →
• Assess and document the lawful rights to collect and use: Which action most directly addresses the threshold →
• All 426 Understanding How to Govern AI Development questions →