Constrain the chatbot's scope so out-of-domain requests | 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 →

Constrain the chatbot's scope so out-of-domain requests: To reduce this unintended downstream harm, which

AIGP Understanding How to Govern AI Deployment and Use Medium

Constraining the chatbot to its intended domain and refusing out-of-scope queries reduces the downstream harm.

The question

A retailer's customer-service chatbot, built only to answer order and shipping questions, is increasingly being asked for medical and legal advice, and it sometimes responds confidently and incorrectly. To reduce this unintended downstream harm, which measure is most appropriate?

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

  1. Constrain the chatbot's scope so out-of-domain requests are refused or routed elsewhere, preventing confident answers outside its intended use. ✓
    Correct because scoping the system to its intended domain and refusing out-of-scope queries directly reduces the unintended downstream harm described.
  2. Expand the chatbot's training so it can also answer medical and legal questions accurately and reliably alongside its existing order-support duties.
    Plausible as improvement, but broadening into regulated advice domains multiplies risk rather than containing the unintended use.
  3. Increase the chatbot's response speed so customers receive their order and shipping answers with noticeably lower waiting times.
    Plausible as UX work, but faster responses do nothing to prevent harmful out-of-scope answers.
  4. Add a cheerful greeting and friendlier tone so the chatbot feels more approachable to customers during support conversations.
    Plausible as design polish, but tone changes do not address the substantive harm of confident wrong advice outside scope.
The trap
Trying to eliminate misuse by expanding the system's capabilities instead of constraining its scope.

How to remember it

Constraining the chatbot to its intended domain and refusing out-of-scope queries reduces the downstream harm.

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.